mirror of https://github.com/github/spec-kit.git synced 2026-08-03 06:26:30 +08:00

Files

Marsel Safin 683bfd00c9 fix: register extensions for the active integration only (#3459 )

* fix: register extensions for the active integration only

extension add registered commands for every detected agent, and
integration upgrade back-filled enabled extensions for non-active
integrations. Maintainer direction on #2948: treat the project as
single-active. Only the active integration gets extension artifacts;
use/switch rescaffold the target when the user selects it.

- extension add now routes through the all-agents pass restricted to
  the active integration (only_agent), keeping detection and
  missing-skills-dir recovery safeguards. Projects without recorded
  init-options fall back to detection-based registration.
- integration upgrade re-registers extensions only when upgrading the
  active integration, reversing the #2886 back-fill for non-active
  targets at maintainer request.

Fixes #2948

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: address review feedback on active-only extension registration

- Restrict the extension-add active-integration fallback to projects
  with no recorded active key at all. A recorded but unsupported key
  (e.g. "generic", deliberately excluded from AGENT_CONFIGS) no longer
  falls back to registering every detected agent.
- Apply the same single-active rule to preset command overrides:
  PresetManager._register_commands now scopes registration to the
  active integration via only_agent.
- Add PresetManager.register_enabled_presets_for_agent, mirroring
  ExtensionManager.register_enabled_extensions_for_agent, and call it
  from integration use/switch/upgrade (active only) alongside the
  existing extension re-registration so presets are rescaffolded on
  activation instead of being written for inactive integrations.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: address second round of review feedback (priority order, fail-closed, docs)

- register_enabled_presets_for_agent now processes presets in reverse
  priority order (lowest-precedence first) so the highest-precedence
  preset is written last and actually wins after `integration use`
  rescaffolds two overlapping preset command overrides. Verified this
  reproduces the previously reported reversed-priority bug and that the
  fix resolves it.
- _register_commands_for_active_agent now checks for the "ai" key's
  presence separately from its value: a missing key still falls back to
  detection-based registration for all agents, but a recorded, malformed
  value (non-string or empty, e.g. [] or null) now fails closed
  (registers nothing) instead of being treated as "no active
  integration" or reaching AGENT_CONFIGS.get() with an unhashable key
  and raising TypeError.
- Updated docs/reference/presets.md and docs/reference/integrations.md
  to describe active-only preset/extension registration and clarify
  that `integration use`/`switch` is the activation point for
  installed extensions and presets, and that `upgrade` only
  re-registers them for the active integration.

Adds regression tests: two enabled presets overriding the same command
with different priorities (priority winner must survive `use`
rescaffolding), and a malformed recorded `ai` value ([]) for
`extension add`.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: address third round of review feedback (multi-integration semantics)

Fixes five deeper active-only registration bugs surfaced by Copilot review
after 2486c08, all in the presets/extensions single-active integration
rule (#2948):

1. presets: _reconcile_composed_commands (run after install/remove)
   bypassed the active-only filter entirely, writing composition-winner
   command files for every detected non-skill agent via
   register_commands_for_non_skill_agents. Added an only_agent param to
   that registrar method (mirroring register_commands_for_all_agents)
   and threaded it through all 5 reconciliation call sites.

2. presets: `integration use copilot` with --skills (ai_skills: true)
   wrote both the static .agent.md command file AND the SKILL.md
   mirror for the same override. Mirrored the extension path's
   ai_skills guard in both _register_commands and the reconciliation
   pass: a command-backed active agent running in skills mode is
   excluded from non-skill command registration.

3. presets: registered_skills was a flat list, so switching between
   two skill-mode agents (e.g. Claude -> Codex) and then removing the
   preset only restored the currently active agent's directory,
   permanently orphaning the other. _unregister_skills now restores
   every existing skill-mode agent directory instead of only the
   active one.

4. extensions: load_init_options() collapses "no file" and "corrupted
   file" into the same {}, so the round-2 fail-closed fix didn't
   actually distinguish them. Added a shared
   resolve_active_agent_for_registration() helper in _init_options.py
   that checks file existence separately from parse success, returning
   a distinct sentinel for "file absent" vs None for "corrupted or
   invalid". extensions/__init__.py now uses this helper.

5. presets: same corruption-collapsing bug in _register_commands's
   active_agent resolution. Now uses the same shared helper as (4).

Adds regression tests for all five: reconciliation active-only
filtering, copilot --skills dual-write prevention, multi-skill-agent
switch+remove, and corrupted init-options fail-closed behavior for
both extension add and preset add. Each test was verified to fail
against the pre-fix code and pass with the fix.

Targeted (883) and full (3923 passed, 109 skipped) suites pass; ruff
check clean.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: address fourth round of review feedback (skill registration provenance)

Replace the "enumerate every skill-mode directory and restore all of them"
approach from the previous round with precise per-agent provenance
tracking, per reviewer feedback that the enumerate-and-restore-everything
design was unsound:

- registered_skills changes from a flat List[str] to Dict[str, List[str]]
  (agent name -> skill names actually written), mirroring the shape
  registered_commands already uses. _register_skills now returns this
  per-agent mapping instead of a bare list, and every call site
  (register_enabled_presets_for_agent, install_from_directory, the
  _reconcile_skills "was this skill previously managed" check) is updated
  to read/merge the new shape. Legacy flat-list registry entries from
  before this change are still readable: writes self-migrate the format,
  and _normalize_registered_skills() handles the transitional read paths.

- _unregister_skills now restores exactly the agent directories recorded
  for a preset instead of guessing at every skill-mode integration that
  happens to exist on disk. This fixes two problems with the old
  enumerate-everything design: (1) it could silently overwrite or delete
  another preset's (or a user's) override in an agent directory the
  current preset never actually touched, and (2) it depended on
  transient per-process integration state (_skills_mode), which is unset
  in a fresh CLI invocation for mode-selectable integrations like Copilot
  --skills, permanently orphaning their overrides after a process
  restart. Registries written before this change (flat list, no agent
  provenance) fall back to best-effort restoration under only the
  currently active agent, matching the pre-existing guarantee level.

- Every directory resolved from persisted provenance is now validated
  through the project's shared symlink/containment guard
  (_ensure_safe_shared_directory) before any file in it is read, written,
  or removed, since restoration may target an agent that isn't currently
  active and its directory can't be assumed safe just because a name was
  recorded for it.

- _tracked_skill_agent_dirs() (the enumeration helper introduced last
  round) is removed; it's superseded by the provenance-based design.

Adds regression tests: a symlinked skills directory is rejected during
removal; removing one preset does not disturb a different preset's
override in another agent's directory; and a Copilot --skills
registration installed, then removed after switching agents in a fresh
PresetManager instance (simulating a new process), is still correctly
restored. Updates existing skill-registration assertions across
test_presets.py and test_integration_claude.py for the new per-agent
registry shape.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: address fifth round of review feedback (symlink presence, rescaffold reconciliation, shared skills dir)

- _init_options.py: resolve_active_agent_for_registration() now treats a
  dangling init-options.json symlink as present (path.is_symlink() check
  alongside path.exists()), since Path.exists() follows symlinks and
  returns False for a broken one. Previously a broken symlink fell back
  to the legacy "no file" path and registered every detected agent
  instead of failing closed.
- presets/__init__.py (register_enabled_presets_for_agent): the
  integration use/switch rescaffold path now collects affected command
  names across all presets processed and runs
  _reconcile_composed_commands/_reconcile_skills once after the loop,
  matching install/remove. Previously rescaffolding wrote each preset's
  raw content directly with no follow-up reconciliation, so a
  project-level override (the highest-priority layer) could be clobbered
  by a lower-precedence preset after switching agents.
- presets/__init__.py (_unregister_skills): multiple integrations can
  share one physical skills directory (agy/codex/zed all resolve to
  .agents/skills). Provenance restoration now groups recorded agent
  entries by resolved directory and restores each physical directory
  exactly once, preferring the currently active agent's renderer when it
  owns that directory (otherwise any recorded owner, chosen
  deterministically). Previously each recorded agent key triggered its
  own restore pass against the same directory, with whichever agent was
  iterated last silently winning regardless of which agent was active.

Adds regression tests for each: a dangling init-options.json symlink
failing closed for both preset resolution and extension add; integration
use rescaffold preserving a project override over a lower-priority
preset; and a codex/agy shared-directory removal restoring the directory
exactly once in the active agent's format.

Targeted (tests/integrations/test_integration_subcommand.py,
tests/test_presets.py, tests/test_extensions.py,
tests/test_extension_skills.py,
tests/integrations/test_integration_opencode.py,
tests/integrations/test_integration_claude.py): 930 passed.
Full suite: 3930 passed, 109 skipped.
ruff check: clean on files touched by this change.

Refs #2948

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: guard skill subdirectories and active-agent scoping in preset reconciliation

Fix 4 issues from round-6 review of the active-only integration
registration work (#2948):

- remove(): removed_cmd_names only collected primary command names from
  registered_commands + manifest aliases, missing commands that were
  only ever registered via skills mode (ai_skills guard returns no
  command names for command-backed integrations in skills mode). This
  skipped reconciliation entirely when removing a higher-priority
  skills-mode preset, causing _unregister_skills() to fall back to
  core/extension content instead of the surviving lower-priority
  preset's override. Now every command template's primary name is
  added to removed_cmd_names unconditionally.

- _reconcile_composed_commands(): the "composed is None" branch (fires
  when no replace-strategy layer remains for a command, e.g. after
  removing a wrap/append preset's base) called unregister_commands()
  across every configured non-skill agent, ignoring only_agent. This
  deleted historical artifacts from integrations that were never active
  for the preset. Now filtered by only_agent like the rest of the file.

- Added _validate_skill_subdir() helper (reusing
  _ensure_safe_shared_directory/_validate_safe_shared_directory from
  shared_infra.py) and applied it at every site that reads or writes an
  individual skill subdirectory (_register_skills,
  _unregister_skills_in_dir, _reconcile_skills' override_skills
  restoration loop). _safe_skills_dir_for_agent only validated the
  parent skills directory; a symlinked leaf subdirectory (e.g.
  .claude/skills/speckit-specify) would slip past that check since
  is_dir()/exists() follow symlinks, letting write_text/rmtree operate
  through it to an arbitrary location outside the project.

Added regression tests: removing a higher-priority skills-only preset
restores the surviving lower-priority preset's content; composed-is-None
unregistration only touches the active agent; symlinked skill subdirectory
rejected on restore; symlinked skill subdirectory rejected on write.

Targeted (934) and full (3934 passed, 109 skipped) test suites and ruff
check pass clean.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: persist command registration before fallible skills phase on rescaffold

Fix remaining round-6 review findings on the active-only integration
registration work (#2948):

- register_enabled_presets_for_agent(): registered_commands and
  registered_skills were merged and persisted together in a single
  registry.update() call after both the commands and skills phases ran.
  If _register_skills() raised, the per-preset try/except swallowed it
  before that update() call was reached, even though _register_commands()
  had already written a real command file to disk. That file became
  untracked, so preset removal could no longer clean it up.
  install_from_directory() already persists registered_commands
  immediately after the commands phase, before starting the independently
  fallible skills phase; rescaffold now does the same.

- test_presets.py: renamed a misleading claude_dir variable (pointing at
  Gemini's command directory) in
  test_composed_none_unregister_respects_active_agent to reuse the
  existing gemini_commands_dir variable already defined earlier in the
  same test.

Added regression test
test_rescaffold_persists_commands_before_fallible_skills_phase:
simulates a skills-phase failure during rescaffold and asserts the
command file already written to disk is still tracked in
registered_commands.

Verified all other round-6 findings (preset active-integration scoping,
preset reconciliation/remove paths, skills-mode switching, override
precedence during rescaffold, skill-subdirectory symlink safety) are
already addressed by prior commits in this branch; re-checked each
against current code before concluding no further change was needed.

Targeted (tests/test_presets.py, tests/test_extensions.py: 689 passed)
and full (3935 passed, 109 skipped) suites and ruff check on changed
files pass clean.

Assisted-by: GitHub Copilot (model: Claude Sonnet 5, autonomous)
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: unregister stale opposite-mode preset artifact on same-agent skills toggle

Fix an Important gap in register_enabled_presets_for_agent() surfaced by
quality review (#2948): toggling ai_skills for the *same already-active*
command-backed agent (e.g. `integration upgrade copilot` after flipping
ai_skills, with copilot staying active throughout) left a stale artifact
from the previous mode behind, violating the command/skill mutual-
exclusion invariant this PR otherwise enforces.

- command -> skills: _register_commands()'s ai_skills guard makes the
  commands phase a no-op, but the previously-written command file (e.g.
  .agent.md) and its registered_commands[agent] entry were never cleaned
  up, so it lingered alongside the newly written SKILL.md.
- skills -> command: _get_skills_dir() stops resolving a skills directory
  once ai_skills is off, making the skills phase a no-op, but the
  previously-written SKILL.md and its registered_skills[agent] entry were
  never cleaned up, so it lingered alongside the newly (re)written command
  file.

register_enabled_presets_for_agent() now resolves once per call whether
agent_name is a command-backed integration (extension != "/SKILL.md") and
the current ai_skills state, then narrowly unregisters the stale opposite-
mode entry for that agent via the existing _unregister_commands /
_unregister_skills helpers before persisting updated tracking — mirroring
the same per-agent, per-preset isolation already used elsewhere in this
method. Native skill-only agents (claude, codex, ...) are unaffected:
they have no command/skill toggle, so registered_commands and
registered_skills legitimately co-exist for them by design. The trailing
reconciliation pass, project-override precedence, and per-preset
partial-failure isolation are all unchanged.

Added red-first regression tests exercising the real install +
register_enabled_presets_for_agent rescaffold path in both toggle
directions:
- test_rescaffold_toggle_command_to_skills_removes_stale_command_file
- test_rescaffold_toggle_skills_to_command_removes_stale_skill_file

Both failed against the prior code (stale artifact persisted / registry
still tracked it) and pass after the fix.

Targeted (tests/test_presets.py, tests/test_extensions.py: 691 passed)
and full (3937 passed, 109 skipped) suites and ruff check on changed
files pass clean.

Assisted-by: GitHub Copilot (model: Claude Sonnet 5, autonomous)
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: migrate legacy flat-list registered_skills on rescaffold even when unchanged

Fix a valid finding from GitHub Copilot's review of HEAD b9d9053 (#2948):
register_enabled_presets_for_agent() normalizes a legacy flat-list
registered_skills value (predating per-agent provenance) to the
{agent_name: [...]} dict shape in memory via _normalize_registered_skills,
but the persistence check only compared the two *normalized* forms. When
the freshly rescaffolded skill names are identical to what the legacy
list already held — the common case, since nothing about the preset or
skill actually changed — that comparison is a no-op and registry.update()
is skipped, leaving the *raw* on-disk value as the un-migrated flat list.

A later switch to a different skill-mode agent and removal then follows
_unregister_skills's legacy best-effort path (restore only the currently
active agent's directory) instead of the per-agent provenance path,
permanently orphaning the first agent's override.

Fix: track the raw (pre-normalization) existing value and force
persistence whenever it's a non-empty list, independent of whether the
normalized content changed. Traced registered_commands for the same
class of bug: its registry value has always been Dict[str, List[str]]
(no legacy flat-list format ever existed for it — the existing
`if not isinstance(existing_commands, dict): existing_commands = {}`
guard is not a lossy migration path), so this fix stays scoped to
registered_skills only.

Added red-first regression test
test_rescaffold_migrates_legacy_flat_list_registered_skills: installs a
preset, overwrites its registry entry with a raw legacy flat list,
rescaffolds the *same* active agent with unchanged skill names, and
asserts the raw registry is migrated to per-agent dict form. Extends the
scenario with a switch to a second skill-mode agent and preset removal
to prove both agents' directories restore cleanly instead of orphaning
the first. Failed against the prior code (raw value stayed a list) and
passes after the fix.

Targeted (tests/test_presets.py, tests/test_extensions.py: 692 passed)
and full (3938 passed, 109 skipped) suites and ruff check on changed
files pass clean.

Assisted-by: GitHub Copilot (model: Claude Sonnet 5, autonomous)
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: reconcile before fallible skills phase, infer legacy skill provenance, and unregister stale extension artifacts on toggle

Three findings from the Copilot review on HEAD b9d9053/3a1e749:

1. `register_enabled_presets_for_agent()` only recorded a preset's command
   names into `affected_cmd_names` (the set later passed to
   `_reconcile_composed_commands`/`_reconcile_skills`) in the loop that ran
   *after* `_register_skills()`, inside the same per-preset `try` block. If
   `_register_skills` raised, the `except` caught it and `continue`d before
   that loop ever ran — so a preset whose commands phase already wrote real
   content to disk never got reconciled against the full priority stack,
   leaving its raw content in place instead of a project override or
   higher-precedence preset's content. Fix: record the manifest's command
   names immediately after the commands phase succeeds and persists, before
   calling the independently fallible `_register_skills()`.

2. The legacy flat-list `registered_skills` migration (added for the
   previous review round) attributed every name in the list to whichever
   agent was currently being (re)activated. If the first operation after
   upgrading from a pre-#2948 registry was a direct switch to a *different*
   skill-mode agent (e.g. a legacy Claude override, then `integration use
   codex` with no intervening Claude rescaffold), the migrated dict only
   recorded `{"codex": [...]}`, permanently losing Claude's actual
   provenance and orphaning its override on later removal. Fix: added
   `_infer_legacy_skill_provenance()`, which probes every configured
   skill-mode agent's directory (via the same safe, symlink-validated
   helpers already used for restore/removal) for a `SKILL.md` whose
   frontmatter records this exact preset as the owner
   (`metadata.source == "preset:<pack_id>"`). A name found under more than
   one directory is attributed to every matching agent (the preset may have
   been active while the user switched between several skill-mode agents
   before provenance tracking existed); names that can't be matched to any
   directory still fall back to the previously-active best-effort
   behaviour. Directory grouping for shared-path aliases (e.g.
   agy/codex/zed all resolving to `.agents/skills`) intentionally does not
   call `.resolve()` on the path, since doing so diverges from
   `project_root`'s own resolution state on platforms where a path
   component is itself a symlink (e.g. macOS's `/var` -> `/private/var`)
   and made every subsequent containment check spuriously fail.

3. `register_enabled_extensions_for_agent()` has the same command/skill
   mutual-exclusion gap the preset path had (fixed in a previous round):
   toggling `ai_skills` for the *same active* agent left the opposite
   mode's artifact behind. Command -> skills left the extension's
   `.agent.md` file and its `registered_commands[agent]` entry in place
   once `skills_mode_active` made the commands phase a no-op. Skills ->
   command left the extension's `SKILL.md` file in place, since an empty
   `_register_extension_skills()` result (because this agent's skills
   directory no longer resolves once `ai_skills` is off) was treated as
   "nothing to register" rather than "this was rendered here before and is
   now stale". This diverges from the preset path in one respect:
   `registered_skills` for extensions has always been a flat list with no
   per-agent provenance (extension skills are only ever rendered for the
   active agent, never per-preset-per-agent tracked), so the fix resolves
   ownership by checking which of the extension's tracked skill names
   still exist as directories under this specific agent's directory before
   removing them — mirroring the same technique `unregister_agent_artifacts`
   already uses for full agent deactivation, but scoped narrowly to firing
   only when a toggle is actually detected (`skills_mode_active` /
   `command_mode_active`), so a same-mode re-run never disturbs
   already-correct artifacts or a user's manual customizations.

Regression tests (all confirmed red before their respective fix, green
after):
- tests/test_presets.py::TestPresetSkills::test_rescaffold_reconciles_override_even_when_skills_phase_fails
- tests/test_presets.py::TestPresetSkills::test_rescaffold_legacy_flat_list_direct_switch_preserves_original_agent
- tests/test_extension_skills.py::TestExtensionSkillRegistration::test_rescaffold_toggle_command_to_skills_removes_stale_extension_command_file
- tests/test_extension_skills.py::TestExtensionSkillRegistration::test_rescaffold_toggle_skills_to_command_removes_stale_extension_skill_file

Verification: tests/test_presets.py + tests/test_extensions.py +
tests/test_extension_skills.py (753 passed), tests/integrations/ (1768
passed, 1 skipped), full suite `pytest tests -q` (3942 passed, 109
skipped), `ruff check` on changed files clean.

Assisted-by: GitHub Copilot (model: Claude Sonnet 5, autonomous)
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: broaden legacy skill provenance inference to command-backed agents

_infer_legacy_skill_provenance() only probed agents whose registrar
config statically declares extension == "/SKILL.md", excluding
command-backed agents (e.g. Copilot) that can also render preset
overrides as SKILL.md files when ai_skills is enabled. A real
preset-owned .github/skills/.../SKILL.md written while Copilot was the
active skills-mode agent was therefore never probed and got
misattributed entirely to whichever agent activated first after the
upgrade, permanently orphaning Copilot's override on later removal.

Broaden the candidate set to every configured integration
(CommandRegistrar.AGENT_CONFIGS), reusing the existing safe-path
helper (_safe_skills_dir_for_agent, itself built on the shared
_get_skills_dir resolver) rather than inventing new path-construction
logic. The existing preset-marker match (metadata.source ==
"preset:<pack_id>") continues to gate every attribution, so
command-mode agents that never rendered this preset's skill are not
falsely attributed.

Add red-first regression tests: a legacy flat-list entry owned by
Copilot in skills mode, switched directly to Claude with no
intervening Copilot rescaffold, now migrates to a per-agent dict
covering both agents, and removal restores both agents' files instead
of orphaning Copilot's override; plus a negative-case test confirming
a command-mode Copilot with no preset-owned skill marker is not
falsely attributed during the same migration.

Assisted-by: GitHub Copilot (model: Claude Sonnet 5, autonomous)
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: preserve extension skill tracking for mirrors in other agent dirs

The skills -> command toggle cleanup in
register_enabled_extensions_for_agent() recomputed the remaining
tracked registered_skills names by checking only the toggling agent's
own skills directory. Since registered_skills is a single flat list
shared across every agent an extension has ever been activated under
(skills are only ever rendered for the active agent, so there is no
per-agent registry key), a name whose mirror still existed under a
*different*, previously-active agent's directory was incorrectly
dropped from tracking as soon as the current agent's own copy was
removed. A later full removal only iterates registered_skills, so the
orphaned mirror under the other agent's directory was never found or
cleaned up.

Add _extension_owned_skill_names(), which re-verifies ownership across
every configured agent's skills directory (deduped by shared path) the
same way the existing _unregister_extension_skills() fallback scan
already does, keeping a name only when a SKILL.md with a matching
metadata.source == "extension:<id>" marker is found somewhere -
read-only, no directory creation, no symlink escape. Use it instead of
re-checking only the toggling agent's own directory when recomputing
what remains tracked after narrow stale-mirror cleanup.

Add a red-first regression test: Auggie is activated in skills mode
first (writing a mirror), then Copilot is activated in skills mode
(writing its own mirror for the same names), then Copilot toggles to
command mode. Before the fix, registered_skills lost both names
entirely even though Auggie's mirrors were untouched on disk; after
the fix tracking is preserved and a subsequent full removal correctly
cleans up Auggie's remaining mirrors too.

Assisted-by: GitHub Copilot (model: Claude Sonnet 5, autonomous)
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: reject symlinked skills-directory escape in extension skill scans

_extension_owned_skill_names() and the fast/fallback paths of its
sibling _unregister_extension_skills() called skills_candidate.resolve()
and then checked children relative to that already-resolved candidate.
If the candidate directory itself (e.g. .gemini/skills) was a symlink
pointing outside the project root, both the resolve() call and the
subsequent containment check silently passed through the symlink
instead of rejecting it:

- _extension_owned_skill_names() would falsely attribute ownership to
  a marker-matching SKILL.md living outside the project.
- _unregister_extension_skills()'s fast path (an explicit skills_dir,
  as passed by the toggle-cleanup call site) and its fallback scan
  (used during full extension removal) would both shutil.rmtree() the
  external directory, deleting unrelated content outside the project.

Fix by validating the candidate directory itself with the existing
_validate_safe_shared_directory() shared-infra helper before any probe
or delete: it rejects a symlink at any path component (walking down
from the project root, including the final component) without ever
resolving through it, and is already used elsewhere in the codebase for
the same class of shared-directory containment check. Unsafe
candidates are skipped/refused rather than followed.

Add red-first security regression tests reproducing each of the three
call sites with a `.gemini/skills` symlink pointing at an external
directory containing a marker-matching SKILL.md and an unrelated
precious_file.txt: provenance inference must not attribute the name,
and both the explicit-skills_dir fast path and the None-skills_dir
fallback scan must leave the external directory and file untouched.
Existing valid shared/deduped directory tests (e.g. agy/amp/codex/zed
sharing .agents/skills) continue to pass, confirming legitimate shared
directories still clean up correctly.

Assisted-by: GitHub Copilot (model: Claude Sonnet 5, autonomous)
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* Fix unscoped extension-skill removal and legacy preset provenance on direct remove

- _unregister_extension_skills(): omitting skills_dir now always triggers
  the full multi-directory fallback scan instead of narrowing to the
  currently active agent's directory. Previously, remove() (the only
  caller that omits skills_dir) would resolve the active agent's dir and
  take the scoped fast path, orphaning a previously-active second agent's
  extension skill mirror during full removal.

- PresetManager.remove(): infer legacy flat-list registered_skills
  provenance (reusing _infer_legacy_skill_provenance from the prior
  rescaffold fix) before invoking _unregister_skills, so a direct
  `preset remove` with no intervening rescaffold/switch also restores
  every previously-active agent's directory instead of only the
  currently active one.

Added regression tests:
- test_remove_while_second_agent_still_in_skills_mode_cleans_up_first_agent_mirror
- test_remove_infers_legacy_flat_list_provenance_without_prior_rescaffold

Assisted-by: GitHub Copilot (model: Claude Sonnet 5, autonomous)
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* Keep unregister_agent_artifacts scoped to its agent when directory is absent

ExtensionManager.unregister_agent_artifacts() converted its resolved
agent_skills_dir to None whenever that directory didn't exist, before
calling _unregister_extension_skills(). After 1d8f9e3, omitting
skills_dir means "genuinely unscoped removal": scan every configured
agent's directory, reserved for ExtensionManager.remove()'s full
project cleanup. Since unregister_agent_artifacts is agent-scoped (used
by switch to clean up the previous integration's artifacts), this
caused it to delete every other agent's live extension skill mirrors
whenever the target agent's own directory happened to be absent, e.g.
unregistering an agent that was never activated.

Fix: always pass the explicit, agent-scoped skills_dir, even when it
doesn't exist on disk, so the fast path is a safe no-op for an absent
directory instead of falling back to the all-agents scan. Registry
reconciliation (dropping removed names from the flat registered_skills
list) now only runs when the agent's directory actually exists, so an
absent directory can't be misread as "these names were removed
everywhere" and wipe tracking for mirrors that still legitimately live
under other agents' directories.

Added regression test:
- test_unregister_agent_artifacts_stays_scoped_when_agent_dir_absent

Assisted-by: GitHub Copilot (model: Claude Sonnet 5, autonomous)
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* Preserve global skill tracking across agents in unregister_agent_artifacts

The present-directory branch of ExtensionManager.unregister_agent_artifacts()
recomputed "remaining" registered_skills only by checking whether each name
still existed under the just-cleaned agent's own directory. registered_skills
is a single flat list shared across every agent an extension was ever
activated under (skills are only ever rendered for the currently active
agent, so there's no per-agent registry key). Repro: auggie and copilot both
have mirrors for the same extension; unregister_agent_artifacts("auggie")
correctly removes auggie's own mirror, sees the names absent from auggie's
(now empty) directory, and stores an empty registered_skills list - even
though copilot's mirror is still live on disk and now untracked. A later full
remove() then reads an empty registry and leaves copilot's mirror orphaned.

Fix: after the agent-scoped cleanup, recompute remaining names with
_extension_owned_skill_names(), which scans every safe, configured agent
skills directory (not just the one just cleaned) and keeps a name only if a
marker-verified SKILL.md for this extension still exists somewhere. This is
the same helper already used for the analogous same-agent toggle-cleanup
case, so no new abstraction was introduced. Explicit per-agent cleanup,
marker ownership verification, and symlink/containment safety are unchanged.

Added regression test:
- test_unregister_agent_artifacts_preserves_tracking_for_other_agent_mirror

Assisted-by: GitHub Copilot (model: Claude Sonnet 5, autonomous)
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* Reconcile every historical agent on preset removal; validate child skill dirs

Fixes 3 findings from the Copilot review on HEAD 31c9b97 (#2948):

1. presets/__init__.py: remove()'s command reconciliation only recreated
   the surviving preset's content for the currently active agent, even
   though the removed preset's registered_commands could span multiple
   historical (now-inactive) agents recorded via prior rescaffolds. Now
   remove() captures every historical agent registered_commands actually
   targeted (before mutation) and passes it as extra_agents through
   _reconcile_composed_commands -> _register_for_non_skill_agents /
   _register_command_from_path -> registrar.register_commands_for_non_
   skill_agents, so the active-only restriction for install/use is
   preserved while post-removal reconciliation restores every touched
   directory.

2. presets/__init__.py: the analogous gap existed for skills. _unregister_
   skills() now returns {skills_dir: renderer_agent} for every directory it
   actually restored, and _reconcile_skills() accepts extra_skills_dirs to
   reconcile each of those directories (via a new apply_to_dir() helper),
   not only the currently active skills directory. _register_skills() gained
   optional target_dir/target_agent overrides (forcing
   create_missing_skills off for non-active directories) so a historical
   directory is only ever restored, never seeded with brand-new skills.

3. extensions/__init__.py: _extension_owned_skill_names() and both the
   fast and fallback paths of _unregister_extension_skills() validated only
   the parent skills_dir for symlink escape, then resolved
   skills_dir / skill_name and checked containment relative to that
   already-resolved parent. A per-skill child that is itself a symlink to
   a different, legitimate skill directory within the same (safe) root
   passed that containment check, so deleting/attributing through the
   symlink name could destroy or misattribute an unrelated skill reached
   only via the alias. All three call sites now run the shared
   _validate_safe_shared_directory() component-wise check against the full
   skills_dir / skill_name path (not just the parent) before any read or
   delete, rejecting a symlinked child outright rather than following it,
   even when its resolved target remains in-bounds.

Regression tests added (all confirmed red against pre-fix code, green
after):
- test_remove_reconciles_command_for_every_historical_agent
- test_remove_reconciles_skill_for_every_historical_agent
- test_extension_owned_skill_names_rejects_symlinked_child_skill_dir
- test_unregister_extension_skills_explicit_dir_rejects_symlinked_child
- test_unregister_extension_skills_fallback_rejects_symlinked_child

Tests: tests/test_presets.py (361), tests/test_extension_skills.py (69),
tests/test_extensions.py (338) all pass; tests/integrations (1768 passed,
1 skipped) pass; full suite 3902 passed / 74 skipped (90 pre-existing,
environment-only git-signing tests deselected — confirmed failing
identically on the pre-change baseline due to local 1Password SSH-agent
signing, unrelated to this change). ruff check clean on all changed files.

Assisted-by: GitHub Copilot (model: Claude Sonnet 5, autonomous)
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* Persist historical reconciliation ownership; defer destructive toggle cleanup; validate registry-provided skill names

Round 11 review findings (5 comments on HEAD ab6c28c), three root causes:

A) Historical-agent reconciliation wrote surviving content to disk but
   discarded the returned per-agent write map, so the preset's own
   registered_commands/registered_skills never learned about directories
   reconciliation restored on its behalf. A later removal of that same
   preset then orphaned those directories. Added
   _merge_pack_registered_commands/_merge_pack_registered_skills and wired
   them into _reconcile_composed_commands and _reconcile_skills's
   apply_to_dir so every actual write is merged back into the winning
   preset's registry metadata.

B) Command<->skills toggle on an already-active agent deleted the old
   artifact before the replacement registration ran, in both
   presets/__init__.py's register_enabled_presets_for_agent and
   extensions/__init__.py's register_enabled_extensions_for_agent. If the
   replacement step raised, both artifacts were lost. Deferred the
   destructive cleanup until after the replacement phase completes
   without raising (register-new-then-remove-old ordering); the mirror
   skills->command direction was already safe since the new command file
   is always registered unconditionally before any cleanup runs.

C) _unregister_skills_in_dir and _infer_legacy_skill_provenance joined a
   registry-provided (untrusted) skill name directly onto a directory
   before any name-shape validation. An absolute in-project name discards
   the intended parent directory entirely (Path's "/" operator drops the
   left side for an absolute right side), letting a corrupted registry
   entry escape the intended skills subtree while still resolving inside
   the project root - passing the existing containment/symlink check.
   Added a centralized _is_safe_registry_skill_name guard (rejecting
   non-strings, empty strings, absolute paths, multi-component paths, and
   "."/".." ) and applied it before every path join derived from
   registry-provided skill names in both functions. Also fixed
   _infer_legacy_skill_provenance's unmatched-name fallback, which
   previously still attributed rejected names to fallback_agent even
   after the loop skipped them.

Added red-first regressions for all three root causes, covering: a
two-preset historical-command-agent survivor scenario, an analogous
skill-agent survivor scenario, injected skills-phase failure during a
preset command->skills toggle and the extension equivalent, a direct
unit test of the new name-safety guard, an absolute-path escape attempt
against _unregister_skills_in_dir, and a false-attribution attempt
against _infer_legacy_skill_provenance.

Tests: tests/test_presets.py (367 passed), tests/test_extension_skills.py
+ tests/test_extensions.py (408 passed), tests/integrations (1768
passed, 1 skipped), full suite tests -q deselecting the pre-existing
1Password-signing-affected tests/extensions/git/test_git_extension.py
(3909 passed, 74 skipped, 90 deselected). ruff check clean on all
changed files.

Assisted-by: GitHub Copilot (model: Claude Sonnet 5, autonomous)
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* Verify replacement actually landed before retiring stale toggle artifacts

The command<->skills toggle cleanup added for #2948 deferred destructive
removal of the old-mode artifact until after the replacement registration
call completed without raising. That was necessary but not sufficient:
none of _register_skills(), _register_commands(),
register_commands_for_agent(), or _register_extension_skills() raise on
a missing source template, a safety-validation skip, or a corrupted
manifest entry — they simply return an empty or partial result. Treating
"did not raise" as "fully replaced" meant a stale artifact could still be
deleted (or its tracking dropped) even though its specific replacement
never actually landed, leaving neither artifact in place for that logical
command/skill.

Fix all four affected toggle directions by checking the replacement
call's actual return value before allowing any destructive step:

- presets command->skills (register_enabled_presets_for_agent): only
  unregister a stale command name once its corresponding skill name
  (via the existing _skill_names_for_command() helper) is confirmed
  present in the skills call's returned names for that agent; the
  remainder stays tracked and on disk.
- presets skills->command (register_enabled_presets_for_agent): only
  unregister a stale skill name once its corresponding command name is
  confirmed present in the commands call's returned names for that
  agent, using the same helper.
- extensions skills->command (register_enabled_extensions_for_agent):
  only remove a skill mirror once the matching command (mapped via the
  existing HookExecutor._skill_name_from_command() helper) is confirmed
  present in register_commands_for_agent's returned names.
- extensions command->skills (register_enabled_extensions_for_agent):
  only remove a deferred stale command once its matching skill name is
  confirmed present in _register_extension_skills()'s returned names.

All four reuse the existing command<->skill name-derivation helpers
rather than inventing new mapping logic. Registry tracking is updated to
retain exactly the unreplaced subset rather than being popped wholesale,
so partially-successful toggles leave correct, minimal tracking behind.

Added 8 new regression tests (4 presets, 4 extensions) covering both the
fully-empty and genuinely-partial result cases for each of the four
toggle directions, using real missing-source-file scenarios (not mocked
return values) to exercise the actual code paths. Confirmed red before
the fix and green after for all 8.

Focused (test_presets.py, test_extension_skills.py, test_extensions.py,
tests/integrations): 2551 passed, 1 skipped.
Full suite (tests, excluding the pre-existing environment-local
1Password-signing git-extension failures): 3917 passed, 74 skipped, 90
deselected.
ruff check: clean.

Assisted-by: GitHub Copilot (model: Claude Sonnet 5, autonomous)
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* Retire alias command groups on toggle; scope preset cleanup to switched-away agent (#2948)

Fixes three current Copilot review findings on HEAD d0d152e:

1. Command->skills toggle cleanup only matched a stale command's own
   name against the returned replacement skill name. Aliases
   (CommandRegistrar tracks and returns primary + alias names flattened
   into one list) never have their own skill rendered -- only the
   primary command's skill is rendered -- so an alias's name could never
   match, leaving its command artifact and tracking behind forever even
   after the primary's replacement landed. Fixed identically in both
   presets (register_enabled_presets_for_agent) and extensions
   (register_enabled_extensions_for_agent): build a primary->alias
   mapping from the manifest, group stale names by primary, and
   retire/keep the whole group together based solely on whether the
   primary's skill replacement actually landed.

2. `integration switch` to a not-yet-installed target unregistered the
   old agent's extension artifacts but had no preset equivalent, so a
   preset's command overrides (including custom preset commands) and
   skill mirrors for the deactivated agent lingered as orphans. Added
   `PresetManager.unregister_agent_artifacts()`, mirroring
   `ExtensionManager.unregister_agent_artifacts()`: scoped strictly to
   the given agent, migrates a legacy flat-list `registered_skills`
   entry via existing on-disk provenance inference before removing
   anything (so other agents' real ownership is preserved rather than
   guessed or dropped), and guards against double-processing an
   artifact through both the commands and skills paths for native
   SKILL.md agents. Wired via a new `_unregister_presets_for_agent()`
   helper into the integration switch command's existing old-agent
   cleanup phase.

Added red-first regression tests:
- tests/test_presets.py: alias-group retire/keep/partial-multi-group
  tests for the command->skills toggle; unregister_agent_artifacts
  scoping tests for commands and legacy-list skill provenance.
- tests/test_extension_skills.py: alias-group retire/keep tests for the
  extension command->skills toggle.
- tests/integrations/test_integration_subcommand.py: end-to-end switch
  test proving a preset's custom command override is cleaned up when
  switching to a not-yet-installed integration, with tracking updated
  correctly and the new agent's registration unaffected.

All new tests confirmed red (AttributeError / orphaned file assertions)
before the fix and green after. Full suite: 3980 passed, 109 skipped.
ruff check clean on all changed files.

Assisted-by: GitHub Copilot (model: Claude Sonnet 5, autonomous)
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: track reconciled extension artifacts

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* Fix native skill preset reconciliation

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* Fix shared native skill cleanup

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* Fix partial preset rescaffold tracking

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* Fix preset agent skill lifecycle

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* Clarify preset removal reconciliation

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix(integrations): address upgrade review feedback

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix(presets): reconcile partial command writes

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: address active artifact cleanup review

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: defer preset skill cleanup to winning command

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: track reconciled and partial preset skills

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: reconcile project overrides to legacy skills

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: harden preset skill writes and rollback

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix(presets): harden legacy skill restoration

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix(presets): preserve non-owned legacy skills

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: validate reconciled skill paths

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix(presets): preserve reconciled skill ownership

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix(presets): clean reconciled agent skills

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: keep legacy cleanup project-local

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix(presets): keep active agent's artifacts in its current mode on remove

A partially failed command<->skills toggle leaves stale tracking
(registered_commands or registered_skills) for the active agent, and
remove() replayed that history regardless of the agent's current mode:

- extra_agents re-admitted the active skills-mode agent into command
  reconciliation, recreating its command file from a surviving lower
  preset even though only_agent excluded it.
- _unregister_skills restored (and _reconcile_skills reapplied) a skill
  artifact for the active command-mode agent instead of deleting the
  preset-owned leftover.

The active agent's participation is now decided exclusively by its
current mode: reconciliation strips it from extra_agents, and removal
routes its stale skills through _delete_agent_preset_skills. Historical
replay still applies to inactive agents only (#2948).

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: filter uninstalled-extension commands in reconciliation; allow active-agent layout change with presets

Two follow-ups to the upstream-main merge:

- Preset reconciliation (_reconcile_composed_commands) now skips
  extension-scoped commands (speckit.<ext>.<cmd>) whose extension is not
  installed, at the single chokepoint every install/remove/rescaffold
  pass funnels through. Registration already refused them, so
  reconciliation could materialize files no registry entry tracks. The
  duplicated per-call-site filters collapse into one
  _extension_installed_for_command helper.

- The #3415 layout-change guard predates this PR's agent-scoped preset
  rescaffold: for the active integration, _register_presets_for_agent
  now re-registers enabled presets in the new layout and retires the
  old layout's stale files, so an active-agent command<->skills toggle
  proceeds and reconciles instead of being rejected. The guard still
  rejects non-active agents (no rescaffold runs for them) and still
  fails closed on an unreadable registry.
  _installed_presets_affecting_agent also understands the per-agent
  dict shape of registered_skills this PR writes, instead of raising
  'malformed'.

Regression tests: rescaffold with an uninstalled extension's command,
CLI-level legacy<->skills toggle with an installed preset (both
directions), secondary-agent rejection, and dict-shaped
registered_skills in the guard helper.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: reject active layout change while a disabled preset owns artifacts

The post-upgrade preset rescaffold iterates enabled presets only, and a
disabled preset's artifacts are deliberately frozen until removal, so an
active-agent command<->skills layout change cannot reconcile them.
_installed_presets_affecting_agent now reports each preset's enabled
state and the guard rejects the migration while any affected preset is
disabled, with re-enable/remove guidance. Enabled presets and non-active
rejection behave as before.

Regression test: disabled preset blocks the toggle untouched; re-enabling
unblocks it and reconciles.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* docs: replace placeholder prefix in two safety comments

Comment-only: spell out why skill deletion is restricted to
project-local directories (flat/legacy provenance cannot prove
home-directory ownership) instead of an undefined placeholder word.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* docs: correct guard-helper docstring to active-only registration model

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: fail closed on non-list values in per-agent preset provenance

A dict-shaped registered_skills/registered_commands entry with a
non-list value (e.g. null) left ownership undecidable but read as "no
artifacts", letting a layout-changing upgrade proceed on a malformed
registry. Validate values are lists and raise
_PresetRegistryUnreadableError otherwise, matching the guard's
fail-closed contract. Unit test covers both fields.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: drop eager extension unregister on layout-changing upgrade

Unregistering the agent's extension artifacts before re-registration
deleted files and registry tracking up front, so a failed or partial
re-registration left the extension with no artifacts at all. Retirement
of each opposite-mode artifact already belongs to
register_enabled_extensions_for_agent's deferred toggle cleanup, which
removes an old artifact only after its replacement is confirmed. Also
keeps disabled extensions consistent with disabled presets: artifacts
stay frozen in place with intact tracking.

Regression test corrupts the installed extension manifest so
re-registration fails, then asserts the old-layout artifacts and their
registry tracking survive the upgrade.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: rescaffold fallback integration after failed switch rollback

When Phase 2 of a switch fails, rollback restores another installed
integration as the default via _set_default_integration but never
re-registered extensions or presets for it. Under active-only
registration the fallback may never have received any artifacts (it
was installed while another integration was active), and Phase 1
already unregistered the outgoing agent's artifacts — leaving the
restored default unusable. Rescaffold both extensions and presets
(best-effort) after the fallback default is successfully restored.

Regression test: secondary codex install with the git extension, a
failing switch to generic, then asserts codex ends up with registered
extension artifacts after rollback.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* docs: explain load-bearing pre-create loop in _reconcile_skills

The per-skill _validate_skill_subdir(create=True) loop looks like dead
code (its result is unused), but it re-creates the tracked skill
subdirectories that _unregister_skills just deleted so
_register_skills's only-overwrite-existing gate passes during a
historical-directory restore. Removing it fails
test_skill_reconciliation_preserves_per_directory_names.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: preserve dashed-description skill tracking

Use the shared frontmatter parser when verifying surviving extension skill mirrors so delimiter substrings cannot hide provenance metadata.

Assisted-by: GitHub Copilot (model: GPT-5.6 Sol, autonomous)

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: skip absent extension skills during reconciliation

Filter extension-scoped commands before skill reconciliation so historical preset tracking and project overrides cannot recreate artifacts for uninstalled extensions.

Assisted-by: GitHub Copilot (model: GPT-5.6 Sol, autonomous)

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* fix: preserve partial native skill cleanup

Coordinate native-skill command cleanup with registered skill coverage per agent and command so partial rescaffolds cannot orphan preset artifacts.

Assisted-by: GitHub Copilot (model: GPT-5.6 Sol, autonomous)

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

---------

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

2026-07-27 14:09:40 -05:00

27 KiB

Raw Blame History

Supported AI Coding Agent Integrations

The Specify CLI supports a wide range of AI coding agents. When you run specify init, the CLI sets up the appropriate command files and directory structures for your chosen AI coding agent — so you can start using Spec-Driven Development immediately, regardless of which tool you prefer.

Supported AI Coding Agents

Agent	Key	Notes
Amp	`amp`
Antigravity (agy)	`agy`	Skills-based integration; skills are installed automatically
Auggie CLI	`auggie`
Claude Code	`claude`	Skills-based integration; installs skills in `.claude/skills`
Cline	`cline`	IDE-based agent
CodeBuddy CLI	`codebuddy`
Codex CLI	`codex`	Skills-based integration; installs skills into `.agents/skills` and invokes them as `$speckit-<command>`
Cursor	`cursor-agent`
Devin for Terminal	`devin`	Skills-based integration; installs skills into `.devin/skills/` and invokes them as `/speckit-<command>`
Factory Droid	`droid`	Skills-based integration; installs skills into `.factory/skills/` and invokes them as `/speckit-<command>`
Firebender	`firebender`	IDE-based agent for Android Studio / IntelliJ
Forge	`forge`
Gemini CLI	`gemini`
GitHub Copilot	`copilot`	Defaults to legacy markdown mode: `.agent.md` command files under `.github/agents/`, companion `.prompt.md` files under `.github/prompts/`, and a `.vscode/settings.json` merge. Pass `--integration-options="--skills"` to scaffold skills as `speckit-<command>/SKILL.md` under `.github/skills/` instead. Legacy markdown mode is deprecated and will stop being the default in a future release.
Goose	`goose`	Uses YAML recipe format in `.goose/recipes/`
Grok Build	`grok`	Skills-based integration; installs skills into `.grok/skills` and invokes them as `/speckit-<command>`
Hermes	`hermes`	Skills-based integration; installs skills globally into `~/.hermes/skills/`
IBM Bob	`bob`	Skills-based integration by default; installs skills as `speckit-<command>/SKILL.md` under `.bob/skills/` and invokes them as `/speckit-<command>`. Pass `--integration-options="--legacy-commands"` to scaffold the deprecated Bob 1.x layout (`.bob/commands/*.md`) instead; that flag will be removed in a future release. Existing legacy installs can migrate with `specify integration upgrade bob --integration-options="--skills"`, which converts them to the skills layout and removes the old command files. If preset overrides are installed, the migration is rejected with an actionable error (preset artifacts cannot yet be reconciled across a layout change) — remove the preset(s), migrate, then reinstall them.
Junie	`junie`
Kilo Code	`kilocode`	Installs commands into `.kilo/commands`; legacy `.kilocode/workflows` installs remain supported as a registration fallback
Kimi Code	`kimi`	Skills-based integration; installs into `.kimi-code/skills/`. `--migrate-legacy` moves old `.kimi/skills/` installs to the new paths
Kiro CLI	`kiro-cli`	Kiro CLI does not substitute `$ARGUMENTS` in file-based prompts, so Spec Kit ships a prose fallback at render time (see Manage prompts and issue #1926). Alias: `--integration kiro`
Lingma	`lingma`	Skills-based integration; skills are installed automatically
Mistral Vibe	`vibe`
Oh My Pi	`omp`	Installs slash commands into `.omp/commands`
opencode	`opencode`
Pi Coding Agent	`pi`	Pi doesn't have MCP support out of the box, so `taskstoissues` won't work as intended. MCP support can be added via extensions
Qoder CLI	`qodercli`
Qwen Code	`qwen`
RovoDev	`rovodev`	Generates `.rovodev/skills/`, prompt wrappers, and `prompts.yml`; runtime dispatch uses `acli rovodev`
SHAI (OVHcloud)	`shai`
Tabnine CLI	`tabnine`
Trae	`trae`	Skills-based integration; skills are installed automatically
ZCode	`zcode`	Skills-based integration; installs skills into `.zcode/skills/` and invokes them as `$speckit-<command>`
Zed	`zed`	Skills-based integration; installs skills into `.agents/skills` and invokes them as `/speckit-<command>`
Generic	`generic`	Bring your own agent — use `--integration generic --integration-options="--commands-dir <path>"` for AI coding agents not listed above

List Available Integrations

specify integration list

Option	Description
`--catalog`	Also browse the catalog (built-in and community). Community integrations that are not built in are only shown here.

Shows the built-in integrations, which one is currently installed, and whether each requires a CLI tool or is IDE-based. When multiple integrations are installed, the list marks the default integration separately from the other installed integrations. The list also shows whether each built-in integration is declared multi-install safe.

Search Available Integrations

specify integration search [query]

Option	Description
`--tag`	Filter by tag
`--author`	Filter by author

Searches the active catalog stack for integrations matching the query. Without a query, lists all available integrations. Must be run inside a Spec Kit project.

Integration Info

specify integration info <integration_id>

Shows catalog details for a single integration, including its description, author, license, tags, source catalog, repository (when available), and whether it is currently active. Must be run inside a Spec Kit project.

Install an Integration

specify integration install <key>

Option	Description
`--script sh\|ps\|py`	Script type: `sh` (bash/zsh), `ps` (PowerShell), or `py` (Python)
`--force`	Opt in to installing alongside integrations that are not declared multi-install safe
`--integration-options`	Integration-specific options (e.g. `--integration-options="--commands-dir .myagent/cmds"`)

Installs the specified integration into the current project. If another integration is already installed, the command only proceeds automatically when all involved integrations are declared multi-install safe. Otherwise, use switch to replace the default integration or pass --force to explicitly opt in to multi-install. If the installation fails partway through, it automatically rolls back to a clean state.

Installing an additional integration does not change the default integration. Use specify integration use <key> to change the default.

Installed extensions and presets are not registered for a non-default integration at install time — they follow the currently active (default) integration only. specify integration use <key> (or switch <key>) is what rescaffolds them for the newly active integration.

Note: All integration management commands require a project already initialized with specify init. To start a new project with a specific agent, use specify init <project> --integration <key> instead.

Version note: Controlled multi-install support was introduced in Spec Kit 0.8.5. If specify integration install <key> says another integration is already installed and only suggests switch or uninstall, check your local CLI with specify version and upgrade it. Running a one-shot command such as uvx --from git+https://github.com/github/spec-kit.git specify ... uses a temporary copy for that command only; it does not update the persistent specify executable on your PATH.

Uninstall an Integration

specify integration uninstall [<key>]

Option	Description
`--force`	Remove files even if they have been modified

Uninstalls the current integration (or the specified one). Spec Kit tracks every file created during install along with a SHA-256 hash of the original content:

Unmodified files are removed automatically.
Modified files (where you've made manual edits) are preserved so your customizations are not lost.
Use --force to remove all integration files regardless of modifications.

Switch to a Different Integration

specify integration switch <key>

Option	Description
`--script sh\|ps\|py`	Script type: `sh` (bash/zsh), `ps` (PowerShell), or `py` (Python)
`--force`	Force removal of modified files during uninstall; when the target is already installed, overwrite managed shared templates while changing the default
`--refresh-shared-infra`	Also overwrite shared infrastructure files even if you customized them (otherwise customizations are preserved)
`--integration-options`	Options for the target integration when it is not already installed

If the target integration is not already installed, equivalent to running uninstall followed by install in a single step. In this mode, --force controls whether modified files from the removed integration are deleted. If the target integration is already installed, switch only changes the default integration, like use; in this mode, --force controls whether managed shared templates are overwritten while the default changes. --integration-options is rejected for already-installed targets because changing integration options requires reinstalling managed files; run upgrade <key> --integration-options ... first, then use <key>. Like use, switch rescaffolds installed extensions and presets for the target integration once it becomes the default.

Use an Installed Integration

specify integration use <key>

Option	Description
`--force`	Overwrite managed shared templates while changing the default

Sets the default integration without uninstalling any other installed integrations. This also refreshes managed shared templates so command references match the new default integration's invocation style. Modified or untracked shared templates are preserved unless --force is used.

use is also the activation point for installed extensions and presets: it re-registers every enabled extension's and preset's command overrides (and skills, for skills-mode agents) for the newly active integration, so artifacts installed while a different integration was active are rescaffolded here rather than at install time.

Upgrade an Integration

specify integration upgrade [<key>]

Option	Description
`--force`	Overwrite files even if they have been modified
`--script sh\|ps\|py`	Script type: `sh` (bash/zsh), `ps` (PowerShell), or `py` (Python)
`--integration-options`	Options for the integration

Reinstalls an installed integration with updated templates and commands (e.g., after upgrading Spec Kit). Defaults to the default integration; if a key is provided, it must be one of the installed integrations. Detects locally modified files and blocks the upgrade unless --force is used. Stale files from the previous install that are no longer needed are removed automatically. Shared templates stay aligned with the default integration even when upgrading a non-default integration.

Enabled extensions and presets are re-registered only when upgrading the currently active (default) integration. A non-default upgrade still refreshes that integration's core commands, but does not re-register its extension or preset layers — use/switch that integration afterward to rescaffold them.

If an upgrade would change an integration between command and skills layouts while preset artifacts are registered for it, the upgrade is rejected before changing files. Remove the affected presets, run the layout-changing upgrade, then reinstall them.

Report Integration Status

specify integration status
specify integration status --json

Reports the current project's integration status without changing files. The status report includes the default integration, installed integrations, multi-install safety, missing managed files, modified managed files, invalid manifest paths, shared Spec Kit infrastructure health, unchecked manifests, and the target integration for default-sensitive shared templates. The JSON form is intended for CI and coding agents that need stable machine-readable status data; it also reports the raw recorded integrations and the integration manifests that were checked when state repair heuristics differ from the recorded file. The command exits 0 when the report status is ok or warning; it exits 1 only when the report status is error. In JSON output, multi_install_safe is null when no installed integration set can be evaluated, such as when the integration state is missing, unreadable, lacks a valid recorded integration list, or records no installed integrations.

Catalog Management

Integration catalogs control where the discovery commands (search and info) look for integrations. Catalogs are checked in priority order.

List Catalogs

specify integration catalog list

Shows the active catalog sources. Project-level sources (when configured) are removable by index; otherwise the active sources are shown as non-removable.

Add a Catalog

specify integration catalog add <url>

Option	Description
`--name <name>`	Optional name for the catalog

Adds a custom catalog URL to the project's .specify/integration-catalogs.yml. The URL must use HTTPS (except http://localhost, http://127.0.0.1, or http://[::1] for local testing).

Remove a Catalog

specify integration catalog remove <index>

Removes a project catalog source by its 0-based index in catalog list.

Catalog Resolution Order

Catalogs are resolved in this order (first match wins):

Environment variable — SPECKIT_INTEGRATION_CATALOG_URL overrides all catalogs
Project config — .specify/integration-catalogs.yml
User config — ~/.specify/integration-catalogs.yml
Built-in defaults — official catalog + community catalog

Integration-Specific Options

Some integrations accept additional options via --integration-options:

Integration	Option	Description
`generic`	`--commands-dir`	Required. Directory for command files
`kimi`	`--migrate-legacy`	Migrate legacy `.kimi/skills/` installs to `.kimi-code/skills/` (including dotted→hyphenated skill naming, e.g. `speckit.xxx` → `speckit-xxx`)
`copilot`	`--skills`	Scaffold commands as agent skills (`speckit-<command>/SKILL.md` under `.github/skills/`, invoked as `/speckit-<command>`) instead of the default legacy markdown mode (`.github/agents/.agent.md` plus `.github/prompts/.prompt.md` and a `.vscode/settings.json` merge). Without this flag, install warns that legacy markdown mode is deprecated.

Example:

specify integration install generic --integration-options="--commands-dir .myagent/cmds"

Scaffold a New Integration

specify integration scaffold <key>

Creates a minimal built-in integration package and a matching test skeleton in the Spec Kit repository, then prints the next steps for wiring it up. Run this command from the Spec Kit repository root. The <key> must be lowercase kebab-case (for example, my-agent).

Option	Description
`--type`	Scaffold template to use: `markdown` (default), `skills`, `toml`, or `yaml`

FAQ

Can I install multiple integrations in the same project?

Yes, but it is intended for team portability rather than the default workflow. Multiple integrations are allowed automatically only when the installed integration and the new integration are declared multi-install safe by Spec Kit. For other combinations, pass --force to acknowledge that multiple agents may see unrelated agent-specific instructions or commands.

Spec Kit tracks one default integration in .specify/integration.json with default_integration, all installed integrations with installed_integrations, per-integration runtime settings with integration_settings, and a dedicated integration_state_schema for future state migrations. The legacy integration field remains as an alias for the default integration.

Which integrations are multi-install safe?

An integration is multi-install safe when it uses a static, unique agent root and command directory, stable command invocation settings, and a separate install manifest whose managed files do not overlap another safe integration. Registry tests enforce those path and manifest invariants. Shared Spec Kit templates remain aligned to the single default integration.

The Command directory column below lists the directory each integration installs its commands or skills into. Context-file targeting is a separate concern from integration multi-install safety: multi_install_safe is an integration declaration about command/skill paths, whereas the optional agent-context extension manages a per-agent context file (for example AGENTS.md or CLAUDE.md) and can even synchronize several anchors at once via its context_files setting. Multiple agents mapping to the same context file is expected there and does not affect whether an integration is multi-install safe; see the agent-context extension for details.

The currently declared multi-install safe integrations are:

Key	Command directory
`auggie`	`.augment/commands`
`claude`	`.claude/skills`
`cline`	`.clinerules/workflows`
`codebuddy`	`.codebuddy/commands`
`codex`	`.agents/skills`
`cursor-agent`	`.cursor/skills`
`droid`	`.factory/skills`
`firebender`	`.firebender/commands`
`gemini`	`.gemini/commands`
`grok`	`.grok/skills`
`junie`	`.junie/commands`
`kilocode`	`.kilo/commands`
`kiro-cli`	`.kiro/prompts`
`lingma`	`.lingma/skills`
`omp`	`.omp/commands`
`pi`	`.pi/prompts`
`qodercli`	`.qoder/commands`
`qwen`	`.qwen/commands`
`shai`	`.shai/commands`
`tabnine`	`.tabnine/agent/commands`
`trae`	`.trae/skills`
`zcode`	`.zcode/skills`

Integrations that share a command directory with another integration, require dynamic install paths such as --commands-dir, or merge shared tool settings are not declared safe by default. They can still be installed alongside another integration with --force.

What happens to my changes when I uninstall or switch?

Files you've modified are preserved automatically. Only unmodified files (matching their original SHA-256 hash) are removed. Use --force to override this.

How do I know which key to use?

Run specify integration list to see all available integrations with their keys, or check the Supported AI Coding Agents table above.

Do I need the AI coding agent installed to use an integration?

CLI-based integrations (like Claude Code, Gemini CLI) require the tool to be installed. IDE-based integrations (like Cursor) work through the IDE itself. Some agents like GitHub Copilot support both IDE and CLI usage. specify integration list shows which type each integration is.

When should I use `upgrade` vs `switch`?

Use upgrade when you've upgraded Spec Kit and want to refresh an installed integration's managed files. Use switch when you want to replace the current default with another integration; if the target is already installed, switch behaves like use.

Do extensions and presets I install apply to every installed integration?

No. Extensions (specify extension add) and presets (specify preset add) register their command overrides for the currently active (default) integration only, even if other integrations are installed. A non-default integration does not receive those artifacts until it becomes the default: specify integration use <key> (or switch <key>) rescaffolds every enabled extension and preset for the newly active integration. specify integration upgrade follows the same rule — it only re-registers extensions and presets when upgrading the active integration.

27 KiB Raw Blame History