fix(extensions): preserve argument-hint in extension Claude SKILL.md (#2916)

Extension-provided commands that declare `argument-hint:` in their
frontmatter had that field dropped from the generated Claude
`.claude/skills/<name>/SKILL.md`, while core template commands keep it.
The extension skill generator built the frontmatter via the shared
build_skill_frontmatter() (name/description/compatibility/metadata only)
and never forwarded argument-hint.

Carry argument-hint from the parsed source command frontmatter into the
skill frontmatter dict before serialization, gated on the integration
exposing inject_argument_hint so only argument-hint-aware agents (Claude)
receive the key and build_skill_frontmatter's shared shape stays unchanged
for every other agent. The value is injected into the dict rather than via
the string-based inject_argument_hint helper, so a folded multi-line
description cannot be split into invalid YAML.

Add regression tests covering a folding description (Claude) and the
non-Claude gate (kimi).

Closes #2903

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
Ali jawwad
2026-06-11 17:28:39 +05:00
committed by GitHub
parent 36ad3cde1b
commit 63a2a17305
2 changed files with 145 additions and 0 deletions

View File

@@ -1053,6 +1053,22 @@ class ExtensionManager:
description,
f"extension:{manifest.id}",
)
# Preserve the command's argument-hint in the generated skill,
# mirroring the core template path (ClaudeIntegration.setup injects
# it for built-in commands). The value is added to the frontmatter
# dict before serialization — rather than via the string-based
# inject_argument_hint helper — so that a folded multi-line
# description cannot be split by the inserted line. Gated on the
# integration exposing inject_argument_hint so only argument-hint
# aware agents receive the key, leaving build_skill_frontmatter's
# shared shape unchanged for every other agent.
argument_hint = frontmatter.get("argument-hint")
if (
argument_hint
and integration is not None
and hasattr(integration, "inject_argument_hint")
):
frontmatter_data["argument-hint"] = str(argument_hint)
frontmatter_text = yaml.safe_dump(frontmatter_data, sort_keys=False).strip()
# Derive a human-friendly title from the command name