Merge pull request #50 from Dongbumlee/Dongbumlee/copilot-sleep-backend

Add Copilot as a SkillOpt-Sleep model backend (CopilotCliBackend) + research-engine MCP plugin

plugins/README.md

@@ -37,7 +37,7 @@ sleep** idea (short-term experience → long-term competence).
 Requirements: Python ≥ 3.10 and the agent's CLI on PATH. All three call the same
 [`run-sleep.sh`](run-sleep.sh) → `python -m skillopt_sleep`, so behaviour is
 identical everywhere. Default backend is `mock` (no API spend); `--backend
-claude|codex` uses your own budget.
+claude|codex|copilot` uses your own budget.
 
 ---
 
@@ -174,7 +174,7 @@ schedule, if you trust it).
 
 | Flag | Default | Meaning |
 |---|---|---|
-| `--backend mock\|claude\|codex` | `mock` | who runs/optimizes (mock = free) |
+| `--backend mock\|claude\|codex\|copilot` | `mock` | who runs/optimizes (mock = free) |
 | `--preferences "..."` | – | your house rules, as a prior |
 | `--gate on\|off` | `on` | strict held-out gate vs. greedy |
 | `--rollouts-k K` | `1` | multi-rollout contrastive reflection |

plugins/copilot/README.md

@@ -45,8 +45,17 @@ Ask Copilot things like *"run the sleep cycle"*, *"what did the last sleep
 propose?"*, *"adopt the staged sleep proposal"*. Copilot calls the MCP tools:
 `sleep_status`, `sleep_dry_run`, `sleep_run`, `sleep_adopt`, `sleep_harvest`.
 
-Each tool takes optional `project`, `backend` (`mock`/`claude`/`codex`), and
-`scope` arguments. Default backend is `mock` (no API spend).
+Each tool takes optional `project`, `backend` (`mock`/`claude`/`codex`/`copilot`), and
+`scope` arguments. Default backend is `mock` (no API spend). The `copilot`
+backend drives the GitHub Copilot CLI (`copilot -p ... --output-format json`)
+and requires the `copilot` CLI to be installed and authenticated.
+
+For speed, the `copilot` backend runs each call against an isolated
+`COPILOT_HOME` with built-in MCP servers and custom instructions disabled, so
+your user MCP servers (including this project's own) are not spawned per call
+(~5x faster). Override with `SKILLOPT_SLEEP_COPILOT_HOME=<dir>`, pick a model
+with `SKILLOPT_SLEEP_COPILOT_MODEL`, or set `SKILLOPT_SLEEP_COPILOT_FULL_ENV=1`
+to use your real Copilot environment instead.
 
 ## Verify the server directly (no Copilot needed)
 

plugins/copilot/mcp_server.py

@@ -45,8 +45,8 @@ _TOOL_SCHEMA = {
     "type": "object",
     "properties": {
         "project": {"type": "string", "description": "Project dir to evolve (default: cwd)."},
-        "backend": {"type": "string", "enum": ["mock", "claude", "codex"],
-                     "description": "mock = no API spend (default); claude/codex = real."},
+        "backend": {"type": "string", "enum": ["mock", "claude", "codex", "copilot"],
+                     "description": "mock = no API spend (default); claude/codex/copilot = real."},
         "scope": {"type": "string", "enum": ["invoked", "all"]},
     },
     "additionalProperties": False,

plugins/copilot/skillopt/README.md

@@ -0,0 +1,98 @@
+# SkillOpt — GitHub Copilot integration
+
+Give **Copilot** (CLI or VS Code) direct access to the **SkillOpt** research
+engine via a tiny **MCP server**. MCP is GitHub's supported way to extend
+Copilot, so this works across Copilot CLI, VS Code, and other MCP clients with
+the same server.
+
+SkillOpt is **validation-gated, text-space skill optimization**: it reflects on
+rollouts, makes bounded edits to a skill, and keeps a change only if it improves
+a held-out validation set. This plugin exposes the repo's training and eval
+entry points (`scripts/train.py`, `scripts/eval_only.py`) as Copilot tools.
+
+> This is the companion to the **SkillOpt-Sleep** plugin (`../mcp_server.py`,
+> `sleep_*` tools). Sleep evolves a *local coding agent* from your past
+> sessions; this server drives the *research* training/eval loops on the
+> benchmark configs in [`../../../configs`](../../../configs).
+
+## What's here
+
+| File | Purpose |
+|---|---|
+| `mcp_server.py` | stdlib-only MCP (stdio) server exposing `skillopt_*` tools |
+| `mcp-config.example.json` | drop-in MCP server config |
+| `copilot-instructions.snippet.md` | paste into `.github/copilot-instructions.md` |
+
+## Install
+
+Requires Python ≥ 3.10. The MCP server itself is pure stdlib, but the tools it
+launches need SkillOpt's runtime deps — install the package first:
+
+```bash
+pip install -e .   # or: pip install -r requirements.txt
+```
+
+1. **Register the MCP server.** Add the server to your Copilot MCP config
+   (Copilot CLI: `~/.copilot/mcp-config.json`; VS Code: your MCP settings).
+   Use `mcp-config.example.json` as a template — set `SKILLOPT_REPO` to this
+   repo's path:
+
+   ```json
+   {
+     "mcpServers": {
+       "skillopt": {
+         "command": "python3",
+         "args": ["/abs/path/SkillOpt/plugins/copilot/skillopt/mcp_server.py"],
+         "env": { "SKILLOPT_REPO": "/abs/path/SkillOpt" }
+       }
+     }
+   }
+   ```
+
+2. **(Optional) Tell Copilot about it.** Append
+   `copilot-instructions.snippet.md` to your repo's
+   `.github/copilot-instructions.md` so Copilot reaches for the tools when the
+   user asks to "optimize a skill" or "train on a benchmark".
+
+## Use
+
+Ask Copilot things like *"what configs can I run?"*, *"optimize the searchqa
+skill"*, or *"evaluate this skill on the dataset"*. Copilot calls the MCP tools:
+`skillopt_list_configs`, `skillopt_train`, `skillopt_eval`.
+
+| Tool | Required args | Notes |
+|---|---|---|
+| `skillopt_list_configs` | — | Lists `configs/**/*.yaml` you can pass as `config`. |
+| `skillopt_train` | `config` | Runs a reflective optimization loop. Long-running; spends budget. |
+| `skillopt_eval` | `config`, `skill` | Evaluates one skill markdown file; no training. |
+
+Common optional args (both train and eval): `env`, `backend`,
+`optimizer_model`, `target_model`, `out_root`, `cfg_options` (space-separated
+`KEY=VALUE` YAML overrides), and `extra_args` (raw passthrough flags for the
+underlying script). `skillopt_train` also accepts `num_epochs`, `batch_size`,
+`seed`, and `use_gate`.
+
+Runs can be very long. The server's subprocess timeout defaults to 6 hours;
+override it with the `SKILLOPT_RUN_TIMEOUT` environment variable (seconds).
+
+## Verify the server directly (no Copilot needed)
+
+```bash
+printf '%s\n' \
+  '{"jsonrpc":"2.0","id":1,"method":"initialize","params":{}}' \
+  '{"jsonrpc":"2.0","id":2,"method":"tools/list"}' \
+  '{"jsonrpc":"2.0","id":3,"method":"tools/call","params":{"name":"skillopt_list_configs","arguments":{}}}' \
+  | SKILLOPT_REPO="$(pwd)" python3 plugins/copilot/skillopt/mcp_server.py
+```
+
+You should see the server info, the three `skillopt_*` tools, and the list of
+benchmark configs.
+
+## Notes / status
+
+- MCP is the stable, official Copilot extension surface, so this is portable
+  across Copilot CLI and IDE from one server.
+- `skillopt_list_configs` is filesystem-only and safe to call anytime;
+  `skillopt_train` / `skillopt_eval` shell out to the repo scripts and require
+  the SkillOpt runtime deps (and, for real backends, model credentials — see
+  [`../../../.env.example`](../../../.env.example)).

plugins/copilot/skillopt/copilot-instructions.snippet.md

@@ -0,0 +1,33 @@
+<!--
+Copy this block into your repo's .github/copilot-instructions.md so Copilot
+knows the SkillOpt research-engine tools exist. (Copilot reads
+copilot-instructions.md automatically as ambient guidance.)
+-->
+
+## SkillOpt (research skill-optimization engine)
+
+This repo exposes the core **SkillOpt** training/eval engine via an MCP server
+(`skillopt`). SkillOpt is validation-gated, text-space skill optimization: it
+reflects on rollouts, makes bounded edits to a skill, and keeps a change only
+if it improves a held-out validation set.
+
+When the user asks to "optimize a skill", "train on <benchmark>", "run
+SkillOpt", "evaluate this skill", or "what configs can I run", use the MCP
+tools:
+
+- `skillopt_list_configs` — list the benchmark YAML configs you can pass as `config`
+- `skillopt_train` — run a reflective skill-optimization loop on a config (long-running; spends API/compute budget)
+- `skillopt_eval` — evaluate a single skill markdown file on a dataset (no training)
+
+Guidance:
+- Always run `skillopt_list_configs` first if you don't already know a valid `config` path.
+- `skillopt_train` and `skillopt_eval` are long-running and consume the user's
+  model backend/budget — confirm the `config`, `backend`, and model choices
+  with the user before launching, and surface the held-out gate result when the
+  run finishes.
+- For one-off YAML overrides use `cfg_options` (e.g. `seed=123 batch_size=40`);
+  for any other underlying flag use `extra_args`.
+
+This is distinct from the **SkillOpt-Sleep** MCP server (`skillopt-sleep`,
+`sleep_*` tools), which evolves a local coding agent from past sessions rather
+than running the research benchmarks.

plugins/copilot/skillopt/mcp-config.example.json

@@ -0,0 +1,11 @@
+{
+  "mcpServers": {
+    "skillopt": {
+      "command": "python3",
+      "args": ["plugins/copilot/skillopt/mcp_server.py"],
+      "env": {
+        "SKILLOPT_REPO": "${workspaceFolder}"
+      }
+    }
+  }
+}

plugins/copilot/skillopt/mcp_server.py

@@ -0,0 +1,229 @@
+#!/usr/bin/env python3
+"""SkillOpt (research engine) — minimal MCP server (stdio, stdlib-only).
+
+Exposes the core SkillOpt skill-optimization engine as MCP tools so any
+MCP-capable client (GitHub Copilot CLI / VS Code, Claude Desktop, etc.) can
+drive it. No third-party deps: speaks JSON-RPC 2.0 over stdio with just the
+handful of MCP methods clients need.
+
+This is the companion to the SkillOpt-Sleep MCP server (``../mcp_server.py``).
+Where Sleep evolves a *local agent* from past sessions, this server drives the
+*research* training/eval loops from this repo (``scripts/train.py`` /
+``scripts/eval_only.py``) against the benchmark configs in ``configs/``.
+
+Tools exposed:
+  - skillopt_list_configs : discover the benchmark YAML configs you can use
+  - skillopt_train        : run a reflective skill-optimization (training) loop
+  - skillopt_eval         : evaluate a single skill on a dataset (no training)
+
+``skillopt_train`` and ``skillopt_eval`` shell out to the repo's entry-point
+scripts and stream back their stdout/stderr. Configure your client to launch:
+  python plugins/copilot/skillopt/mcp_server.py
+"""
+from __future__ import annotations
+
+import glob
+import json
+import os
+import subprocess
+import sys
+
+# Repo root: three levels up from plugins/copilot/skillopt/mcp_server.py
+REPO_ROOT = os.environ.get("SKILLOPT_REPO") or os.path.abspath(
+    os.path.join(os.path.dirname(__file__), "..", "..", "..")
+)
+PROTOCOL_VERSION = "2024-11-05"
+
+# Training/eval runs are long; give the engine plenty of headroom.
+RUN_TIMEOUT_SECONDS = int(os.environ.get("SKILLOPT_RUN_TIMEOUT", "21600"))  # 6h
+
+
+def _list_configs() -> str:
+    """List the benchmark configs available under configs/ (filesystem only)."""
+    pattern = os.path.join(REPO_ROOT, "configs", "**", "*.yaml")
+    paths = sorted(glob.glob(pattern, recursive=True))
+    if not paths:
+        return f"[no configs found under {os.path.join(REPO_ROOT, 'configs')}]"
+    rels = [os.path.relpath(p, REPO_ROOT).replace(os.sep, "/") for p in paths]
+    lines = ["Available SkillOpt configs (pass as `config`):", ""]
+    lines += [f"  - {r}" for r in rels]
+    return "\n".join(lines)
+
+
+def _run_script(script_rel: str, args: dict, *, required: tuple[str, ...] = ()) -> str:
+    """Shell out to a repo entry-point script, mapping args -> --flags."""
+    for key in required:
+        if not args.get(key):
+            return f"[error] missing required argument: {key}"
+
+    py = sys.executable or "python3"
+    cmd = [py, os.path.join("scripts", script_rel)]
+
+    # Ordered flags that the train/eval scripts accept directly.
+    flag_args = (
+        "config", "skill", "split", "env", "backend",
+        "optimizer_model", "target_model", "out_root",
+        "num_epochs", "batch_size", "seed", "use_gate",
+    )
+    for key in flag_args:
+        val = args.get(key)
+        if val is None or val == "":
+            continue
+        cmd += [f"--{key}", str(val)]
+
+    # cfg-options: arbitrary KEY=VALUE YAML overrides (nargs="+").
+    cfg_options = args.get("cfg_options")
+    if cfg_options:
+        if isinstance(cfg_options, str):
+            cfg_options = cfg_options.split()
+        cmd += ["--cfg-options", *[str(x) for x in cfg_options]]
+
+    # extra_args: raw passthrough for any other train/eval flag.
+    extra = args.get("extra_args")
+    if extra:
+        if isinstance(extra, str):
+            extra = extra.split()
+        cmd += [str(x) for x in extra]
+
+    try:
+        proc = subprocess.run(
+            cmd, cwd=REPO_ROOT, capture_output=True, text=True,
+            timeout=RUN_TIMEOUT_SECONDS,
+        )
+    except subprocess.TimeoutExpired:
+        return f"[error] run exceeded {RUN_TIMEOUT_SECONDS}s timeout: {' '.join(cmd)}"
+    except Exception as e:  # noqa: BLE001
+        return f"[error] failed to run script: {e}"
+    out = (proc.stdout or "").strip()
+    err = (proc.stderr or "").strip()
+    body = out + (("\n[stderr]\n" + err) if err else "")
+    return body or f"[done] exit code {proc.returncode}, no output"
+
+
+TOOLS = [
+    {
+        "name": "skillopt_list_configs",
+        "description": "List the benchmark YAML configs under configs/ that can be passed as `config` to train/eval.",
+    },
+    {
+        "name": "skillopt_train",
+        "description": "Run a SkillOpt reflective skill-optimization (training) loop on a benchmark config. Long-running; uses your model backend/budget.",
+    },
+    {
+        "name": "skillopt_eval",
+        "description": "Evaluate a single skill markdown file on a dataset without training (scripts/eval_only.py).",
+    },
+]
+_BY_NAME = {t["name"]: t for t in TOOLS}
+
+_NO_ARGS_SCHEMA = {"type": "object", "properties": {}, "additionalProperties": False}
+
+_COMMON_PROPS = {
+    "config": {"type": "string",
+               "description": "Path to a benchmark YAML config (e.g. configs/searchqa/default.yaml). See skillopt_list_configs."},
+    "env": {"type": "string", "description": "Override the environment/adapter name (e.g. searchqa, alfworld)."},
+    "backend": {"type": "string", "description": "Model backend (e.g. azure_openai, claude, codex, qwen, minimax)."},
+    "optimizer_model": {"type": "string", "description": "Model used for reflection/skill rewriting (the optimizer)."},
+    "target_model": {"type": "string", "description": "Model used to execute tasks (the target)."},
+    "out_root": {"type": "string", "description": "Output directory root for run artifacts."},
+    "cfg_options": {"type": "string", "description": "Space-separated YAML overrides, e.g. 'seed=123 batch_size=40'."},
+    "extra_args": {"type": "string", "description": "Raw passthrough flags for the underlying script, e.g. '--workers 8 --max_turns 30'."},
+}
+
+_TRAIN_SCHEMA = {
+    "type": "object",
+    "properties": {
+        **_COMMON_PROPS,
+        "num_epochs": {"type": "integer", "description": "Number of optimization epochs."},
+        "batch_size": {"type": "integer", "description": "Tasks per optimization step."},
+        "seed": {"type": "integer", "description": "Random seed."},
+        "use_gate": {"type": "string", "enum": ["true", "false"],
+                     "description": "Whether to keep the held-out validation gate on (default on)."},
+    },
+    "required": ["config"],
+    "additionalProperties": False,
+}
+
+_EVAL_SCHEMA = {
+    "type": "object",
+    "properties": {
+        **_COMMON_PROPS,
+        "skill": {"type": "string", "description": "Path to the skill markdown file to evaluate."},
+        "split": {"type": "string", "description": "Dataset split to evaluate (default: all)."},
+    },
+    "required": ["config", "skill"],
+    "additionalProperties": False,
+}
+
+_SCHEMA_BY_NAME = {
+    "skillopt_list_configs": _NO_ARGS_SCHEMA,
+    "skillopt_train": _TRAIN_SCHEMA,
+    "skillopt_eval": _EVAL_SCHEMA,
+}
+
+
+def _result(id_, result):
+    return {"jsonrpc": "2.0", "id": id_, "result": result}
+
+
+def _error(id_, code, message):
+    return {"jsonrpc": "2.0", "id": id_, "error": {"code": code, "message": message}}
+
+
+def _dispatch(name: str, args: dict) -> str:
+    if name == "skillopt_list_configs":
+        return _list_configs()
+    if name == "skillopt_train":
+        return _run_script("train.py", args, required=("config",))
+    if name == "skillopt_eval":
+        return _run_script("eval_only.py", args, required=("config", "skill"))
+    return f"[error] unknown tool: {name}"
+
+
+def handle(req: dict):
+    method = req.get("method")
+    id_ = req.get("id")
+    if method == "initialize":
+        return _result(id_, {
+            "protocolVersion": PROTOCOL_VERSION,
+            "capabilities": {"tools": {}},
+            "serverInfo": {"name": "skillopt", "version": "0.1.0"},
+        })
+    if method in ("notifications/initialized", "initialized"):
+        return None  # notification, no response
+    if method == "tools/list":
+        return _result(id_, {"tools": [
+            {"name": t["name"], "description": t["description"],
+             "inputSchema": _SCHEMA_BY_NAME[t["name"]]}
+            for t in TOOLS
+        ]})
+    if method == "tools/call":
+        params = req.get("params") or {}
+        name = params.get("name")
+        if name not in _BY_NAME:
+            return _error(id_, -32602, f"unknown tool: {name}")
+        text = _dispatch(name, params.get("arguments") or {})
+        return _result(id_, {"content": [{"type": "text", "text": text}]})
+    if method == "ping":
+        return _result(id_, {})
+    return _error(id_, -32601, f"method not found: {method}")
+
+
+def main() -> int:
+    for line in sys.stdin:
+        line = line.strip()
+        if not line:
+            continue
+        try:
+            req = json.loads(line)
+        except Exception:
+            continue
+        resp = handle(req)
+        if resp is not None:
+            sys.stdout.write(json.dumps(resp) + "\n")
+            sys.stdout.flush()
+    return 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())

skillopt_sleep/__main__.py

@@ -9,7 +9,7 @@
 Common flags:
     --project PATH      project to evolve (default: cwd)
     --scope all|invoked harvest scope (default: invoked)
-    --backend mock|claude|codex
+    --backend mock|claude|codex|copilot
     --source claude|codex|auto
     --model NAME
     --lookback-hours N
@@ -36,7 +36,7 @@ from skillopt_sleep.state import SleepState
 def _add_common(p: argparse.ArgumentParser) -> None:
     p.add_argument("--project", default="")
     p.add_argument("--scope", default="", choices=["", "all", "invoked"])
-    p.add_argument("--backend", default="", choices=["", "mock", "claude", "codex"])
+    p.add_argument("--backend", default="", choices=["", "mock", "claude", "codex", "copilot"])
     p.add_argument("--model", default="")
     p.add_argument("--codex-path", default="", help="path to the real @openai/codex binary")
     p.add_argument("--claude-home", default="", help="override ~/.claude (also isolates state)")

skillopt_sleep/backend.py

@@ -24,6 +24,7 @@ import json
 import os
 import re
 import subprocess
+import tempfile
 from typing import Any, Dict, List, Optional, Tuple
 
 from skillopt_sleep.types import EditRecord, ReplayResult, TaskRecord
@@ -788,6 +789,218 @@ class CodexCliBackend(CliBackend):
             except Exception:
                 pass
 
+def resolve_copilot_path(explicit: str = "") -> str:
+    """Find the GitHub Copilot CLI (`copilot`) binary."""
+    if explicit:
+        return explicit
+    env = os.environ.get("SKILLOPT_SLEEP_COPILOT_PATH")
+    if env:
+        return env
+    import shutil
+    found = shutil.which("copilot")
+    return found or "copilot"
+
+
+class CopilotCliBackend(CliBackend):
+    """Drives the GitHub Copilot CLI in non-interactive mode.
+
+    Uses ``copilot -p <prompt> --output-format json`` and parses the emitted
+    JSONL event stream, returning the concatenated ``assistant.message``
+    content. The plain-text / ``--silent`` modes do not reliably stream the
+    response to stdout on all platforms, so JSONL is used for robust capture.
+
+    The call runs in a clean temp cwd with streaming disabled and tools allowed
+    (so non-interactive mode never blocks on a permission prompt); ``_call``'s
+    prompts ask for final-answer text only, so no tool use is expected there,
+    while ``attempt_with_tools`` exposes real, cross-platform callable shims in
+    the working directory for honest tool-call detection.
+
+    Startup overhead is minimised: each invocation points ``COPILOT_HOME`` at a
+    dedicated, isolated config dir (no user ``mcp-config.json``, so the user's
+    MCP servers — including this project's own — are NOT spawned, avoiding a
+    slow recursive launch), and built-in MCP servers / custom instructions are
+    disabled. Auth is read from the OS credential store / token env vars, which
+    live outside ``COPILOT_HOME``, so isolation does not break authentication.
+    Set ``SKILLOPT_SLEEP_COPILOT_HOME`` to override the isolated home, or set it
+    empty / ``SKILLOPT_SLEEP_COPILOT_FULL_ENV=1`` to use the user's real
+    environment instead.
+    """
+
+    name = "copilot"
+
+    def __init__(self, model: str = "", copilot_path: str = "", timeout: int = 240) -> None:
+        super().__init__(model=model or os.environ.get("SKILLOPT_SLEEP_COPILOT_MODEL", ""),
+                         timeout=timeout)
+        self.copilot_path = resolve_copilot_path(copilot_path)
+        self.full_env = os.environ.get("SKILLOPT_SLEEP_COPILOT_FULL_ENV", "") == "1"
+        # Stable isolated home so first-run setup is cached across calls.
+        if self.full_env:
+            self.copilot_home = ""
+        else:
+            self.copilot_home = os.environ.get("SKILLOPT_SLEEP_COPILOT_HOME") or os.path.join(
+                tempfile.gettempdir(), "skillopt_sleep_copilot_home"
+            )
+            try:
+                os.makedirs(self.copilot_home, exist_ok=True)
+            except Exception:
+                self.copilot_home = ""
+
+    def _call(self, prompt: str, *, max_tokens: int = 1024) -> str:
+        clean_cwd = tempfile.mkdtemp(prefix="skillopt_sleep_copilot_")
+        cmd = [
+            self.copilot_path, "-p", prompt,
+            "--output-format", "json",
+            "--stream", "off",
+            "--no-color",
+            "--log-level", "none",
+            "--allow-all-tools",
+            "-C", clean_cwd,
+        ]
+        if not self.full_env:
+            # Drop unneeded startup work: no built-in (github) MCP server and no
+            # AGENTS.md / custom-instruction loading. With an isolated home that
+            # has no mcp-config.json, no user MCP servers spawn either.
+            cmd += ["--disable-builtin-mcps", "--no-custom-instructions"]
+        if self.model:
+            cmd += ["--model", self.model]
+        env = os.environ.copy()
+        if self.copilot_home:
+            env["COPILOT_HOME"] = self.copilot_home
+        try:
+            proc = subprocess.run(
+                cmd, capture_output=True, text=True, timeout=self.timeout, cwd=clean_cwd,
+                encoding="utf-8", errors="replace", env=env,
+            )
+        except Exception:
+            return ""
+        finally:
+            try:
+                import shutil
+                shutil.rmtree(clean_cwd, ignore_errors=True)
+            except Exception:
+                pass
+        return self._parse_jsonl_response(proc.stdout or "")
+
+    @staticmethod
+    def _parse_jsonl_response(raw: str) -> str:
+        parts: List[str] = []
+        for line in raw.splitlines():
+            line = line.strip()
+            if not line or not line.startswith("{"):
+                continue
+            try:
+                obj = json.loads(line)
+            except Exception:
+                continue
+            if obj.get("type") == "assistant.message":
+                content = (obj.get("data") or {}).get("content")
+                if isinstance(content, str) and content:
+                    parts.append(content)
+        return "\n".join(parts).strip()
+
+    def attempt_with_tools(self, task, skill, memory, tools):
+        # Expose REAL, callable tool shims in the working directory so the
+        # gbrain quick-answerer judge (tool_called=search) is validated
+        # honestly: we detect each call from the shim's log, not from a
+        # self-reported marker. The Copilot CLI is the Windows-validated
+        # backend, so the shims must be cross-platform — a bash `#!/usr/bin/env
+        # bash` + chmod shim does NOT execute via `./tool` under PowerShell/cmd,
+        # so on Windows we emit a `.cmd` batch shim instead.
+        import shutil
+        import stat
+        work = tempfile.mkdtemp(prefix="skillopt_sleep_copilottools_")
+        calllog = os.path.join(work, "_tool_calls.log")
+        tool_names = tools or ["search"]
+        is_windows = os.name == "nt"
+        try:
+            for tname in tool_names:
+                if is_windows:
+                    shim = os.path.join(work, f"{tname}.cmd")
+                    with open(shim, "w") as f:
+                        # `%~n0` is the script's own base name (the tool name);
+                        # writing it keeps the calllog line == tool name so the
+                        # honest-detection match below works unchanged.
+                        f.write(
+                            "@echo off\n"
+                            f'echo %~n0>>"{calllog}"\n'
+                            "echo (search results: 3 relevant notes found; use them to answer)\n"
+                        )
+                else:
+                    shim = os.path.join(work, tname)
+                    with open(shim, "w") as f:
+                        f.write(
+                            "#!/usr/bin/env bash\n"
+                            f'echo "{tname}" >> "{calllog}"\n'
+                            'echo "(search results: 3 relevant notes found; use them to answer)"\n'
+                        )
+                    os.chmod(shim, os.stat(shim).st_mode | stat.S_IEXEC | stat.S_IXGRP | stat.S_IXOTH)
+            if is_windows:
+                tool_hint = (
+                    "You have shell tools available in the current directory: "
+                    + ", ".join(f"{t}.cmd" for t in tool_names)
+                    + " (each callable as `" + tool_names[0] + "` or `.\\"
+                    + tool_names[0] + "`). When the skill says to look something "
+                    "up or search before answering, you MUST actually run the "
+                    "tool (e.g. `" + tool_names[0] + " \"query\"`) before giving "
+                    "your final answer."
+                )
+            else:
+                tool_hint = (
+                    "You have shell tools available in the current directory: "
+                    + ", ".join(f"./{t}" for t in tool_names)
+                    + ". When the skill says to look something up or search before "
+                    "answering, you MUST actually run the tool (e.g. `./search \"query\"`) "
+                    "before giving your final answer."
+                )
+            prompt = (
+                "You are completing a task. Apply the skill and memory rules EXACTLY, "
+                "including any rule about searching/looking up before answering. "
+                "Treat a 'Learned preferences' block as HARD CONSTRAINTS that override "
+                "earlier conflicting skill text.\n\n"
+                f"{tool_hint}\n\n"
+                f"# Skill\n{skill or '(none)'}\n\n# Memory\n{memory or '(none)'}\n\n"
+                f"# Task\n{task.intent}\n\n{task.context_excerpt}\n\n"
+                "Return ONLY the final answer text."
+            )
+            cmd = [
+                self.copilot_path, "-p", prompt,
+                "--output-format", "json",
+                "--stream", "off",
+                "--no-color",
+                "--log-level", "none",
+                "--allow-all-tools",
+                "-C", work,
+            ]
+            if not self.full_env:
+                cmd += ["--disable-builtin-mcps", "--no-custom-instructions"]
+            if self.model:
+                cmd += ["--model", self.model]
+            env = os.environ.copy()
+            if self.copilot_home:
+                env["COPILOT_HOME"] = self.copilot_home
+            resp = ""
+            try:
+                proc = subprocess.run(
+                    cmd, capture_output=True, text=True, encoding="utf-8",
+                    errors="replace", timeout=self.timeout, cwd=work, env=env,
+                )
+                resp = self._parse_jsonl_response(proc.stdout or "")
+            except Exception:
+                resp = ""
+            self._tokens += len(prompt) // 4 + len(resp) // 4
+            called: List[str] = []
+            if os.path.exists(calllog):
+                with open(calllog) as f:
+                    logged = {ln.strip() for ln in f if ln.strip()}
+                called = [t for t in tool_names if t in logged]
+            return resp, called
+        finally:
+            try:
+                shutil.rmtree(work, ignore_errors=True)
+            except Exception:
+                pass
+
+
 class DualBackend(Backend):
     """Route operations to two backends, à la SkillOpt's target vs optimizer.
 
@@ -1036,6 +1249,8 @@ def get_backend(
     if n in {"azure-responses", "azure_responses", "aoai-responses", "responses"}:
         eps = [e.strip() for e in azure_endpoint.split(",") if e.strip()] or None
         return AzureResponsesBackend(deployment=model, endpoints=eps)
+    if n in {"copilot", "github_copilot", "copilot_cli", "gh_copilot"}:
+        return CopilotCliBackend(model=model)
     return MockBackend()
 
 

skillopt_sleep/config.py

@@ -36,7 +36,7 @@ DEFAULTS: Dict[str, Any] = {
     "val_fraction": 0.34,         # real tasks reserved to gate updates
     "test_fraction": 0.0,         # real tasks reserved as the final held-out measure
     # ── optimizer ──────────────────────────────────────────────────────────
-    "backend": "mock",            # "mock" | "claude" | "codex"
+    "backend": "mock",            # "mock" | "claude" | "codex" | "copilot"
     "model": "",                  # backend-specific; "" => backend default
     "gate_mode": "on",            # "on" (validation-gated) | "off" (greedy, no hard filter)
     "codex_path": "",             # "" => auto-detect the real @openai/codex binary

skillopt_sleep/experiments/run_experiment.py

@@ -134,7 +134,7 @@ def main(argv=None) -> int:
     ap = argparse.ArgumentParser(description="SkillOpt-Sleep validation experiment")
     ap.add_argument("--persona", default="researcher", choices=list(PERSONAS.keys()))
     ap.add_argument("--nights", type=int, default=4)
-    ap.add_argument("--backend", default="mock", choices=["mock", "claude", "codex"])
+    ap.add_argument("--backend", default="mock", choices=["mock", "claude", "codex", "copilot"])
     ap.add_argument("--model", default="", help="backend model override")
     ap.add_argument("--codex-path", default="", help="path to the real @openai/codex binary")
     ap.add_argument("--edit-budget", type=int, default=4)

tests/test_sleep_engine.py

@@ -509,5 +509,124 @@ class TestFullCycleAndAdopt(unittest.TestCase):
                 self.assertIn("answer", f.read().lower())
 
 
+class TestCopilotBackend(unittest.TestCase):
+    """Pure-logic tests for CopilotCliBackend — no `copilot` CLI required."""
+
+    def test_alias_resolution(self):
+        from skillopt_sleep.backend import CopilotCliBackend, get_backend
+        for name in ("copilot", "github_copilot", "copilot_cli", "gh_copilot"):
+            self.assertIsInstance(get_backend(name), CopilotCliBackend, name)
+
+    def test_parse_jsonl_concatenates_assistant_messages(self):
+        from skillopt_sleep.backend import CopilotCliBackend
+        raw = "\n".join([
+            '{"type":"session.info","data":{}}',
+            '{"type":"assistant.message","data":{"content":"hello"}}',
+            'not-json-noise',
+            '{"type":"user.message","data":{"content":"ignored"}}',
+            '{"type":"assistant.message","data":{"content":"world"}}',
+        ])
+        self.assertEqual(CopilotCliBackend._parse_jsonl_response(raw), "hello\nworld")
+
+    def test_parse_jsonl_ignores_non_assistant_and_blank(self):
+        from skillopt_sleep.backend import CopilotCliBackend
+        self.assertEqual(CopilotCliBackend._parse_jsonl_response(""), "")
+        self.assertEqual(
+            CopilotCliBackend._parse_jsonl_response('{"type":"result","data":{"content":"x"}}'),
+            "",
+        )
+        # assistant.message with empty/missing content contributes nothing
+        self.assertEqual(
+            CopilotCliBackend._parse_jsonl_response(
+                '{"type":"assistant.message","data":{"content":""}}\n'
+                '{"type":"assistant.message","data":{}}'
+            ),
+            "",
+        )
+
+    def test_isolated_home_by_default(self):
+        from skillopt_sleep.backend import CopilotCliBackend
+        be = CopilotCliBackend()
+        self.assertFalse(be.full_env)
+        self.assertTrue(be.copilot_home)  # an isolated COPILOT_HOME is set
+
+    def test_full_env_opt_out(self):
+        from skillopt_sleep.backend import CopilotCliBackend
+        prev = os.environ.get("SKILLOPT_SLEEP_COPILOT_FULL_ENV")
+        os.environ["SKILLOPT_SLEEP_COPILOT_FULL_ENV"] = "1"
+        try:
+            be = CopilotCliBackend()
+            self.assertTrue(be.full_env)
+            self.assertEqual(be.copilot_home, "")  # real user environment used
+        finally:
+            if prev is None:
+                os.environ.pop("SKILLOPT_SLEEP_COPILOT_FULL_ENV", None)
+            else:
+                os.environ["SKILLOPT_SLEEP_COPILOT_FULL_ENV"] = prev
+
+    def test_home_override_env(self):
+        from skillopt_sleep.backend import CopilotCliBackend
+        with tempfile.TemporaryDirectory() as d:
+            target = os.path.join(d, "myhome")
+            prev = os.environ.get("SKILLOPT_SLEEP_COPILOT_HOME")
+            os.environ["SKILLOPT_SLEEP_COPILOT_HOME"] = target
+            try:
+                be = CopilotCliBackend()
+                self.assertEqual(be.copilot_home, target)
+                self.assertTrue(os.path.isdir(target))  # created on init
+            finally:
+                if prev is None:
+                    os.environ.pop("SKILLOPT_SLEEP_COPILOT_HOME", None)
+                else:
+                    os.environ["SKILLOPT_SLEEP_COPILOT_HOME"] = prev
+
+    def test_attempt_with_tools_honest_detection(self):
+        # End-to-end (no real CLI): a tiny per-OS stub stands in for `copilot`.
+        # It runs the local `search` shim the backend writes into its work dir
+        # (so the calllog is written — honest detection) then prints one JSONL
+        # assistant.message. Proves both the JSONL parse and that the tool call
+        # is detected from the shim's log, not from a self-reported marker.
+        import shutil
+        import stat
+
+        from skillopt_sleep.backend import CopilotCliBackend
+
+        stub_dir = tempfile.mkdtemp(prefix="skillopt_sleep_stub_")
+        try:
+            if os.name == "nt":
+                stub = os.path.join(stub_dir, "copilot.cmd")
+                with open(stub, "w") as f:
+                    # The backend writes `search.cmd`; run it (explicit `.\` so
+                    # cmd's `call` resolves it from the cwd reliably) so the
+                    # calllog is populated, then emit the JSONL line. None of
+                    # `{ } " :` need escaping in batch echo (no > < | & ^ %).
+                    f.write(
+                        "@echo off\n"
+                        'call .\\search.cmd "q" >nul 2>&1\n'
+                        'echo {"type":"assistant.message","data":{"content":"Paris"}}\n'
+                    )
+            else:
+                stub = os.path.join(stub_dir, "copilot")
+                with open(stub, "w") as f:
+                    f.write(
+                        "#!/usr/bin/env bash\n"
+                        './search "q" >/dev/null 2>&1\n'
+                        "echo '{\"type\":\"assistant.message\",\"data\":{\"content\":\"Paris\"}}'\n"
+                    )
+                os.chmod(
+                    stub,
+                    os.stat(stub).st_mode | stat.S_IEXEC | stat.S_IXGRP | stat.S_IXOTH,
+                )
+
+            be = CopilotCliBackend(copilot_path=stub, timeout=60)
+            task = TaskRecord(id="t1", project="p", intent="What is the capital of France?")
+            resp, called = be.attempt_with_tools(task, skill="", memory="", tools=["search"])
+
+            self.assertEqual(resp, "Paris")  # JSONL parsed via _parse_jsonl_response
+            self.assertEqual(called, ["search"])  # shim ran; detected from calllog
+        finally:
+            shutil.rmtree(stub_dir, ignore_errors=True)
+
+
 if __name__ == "__main__":
     unittest.main(verbosity=2)