remy-changelog

name	remy-changelog
description	Generate a temporary change log recording code modifications. Paired with /remy-audit: generate log, rewind AI context (preserving code changes), then run blind audit.

Change Log Generation Protocol

You must generate a change log file at .claude/temp_log/_temp_{task_id}_{timestamp}.md. All content MUST be in the language specified by the REMY_LANG environment variable (zh-CN → Chinese, en → English).

External Files

Path Convention: All paths below are relative to ~/.claude/. Use Read("~/.claude/skills/remy-changelog/...") to access them — they are NOT in the project working directory.

File	Purpose
skills/remy-changelog/output_schema.json	Context dict structure definition. Populate all required fields.
skills/remy-changelog/render.py	Template rendering helper. Uses Jinja2 when available, falls back to str formatting.
skills/remy-changelog/templates/changelog.md.j2	Jinja2 template for the change log output.

Optional Dependency: Jinja2

render.py attempts import jinja2. If unavailable, all templates are rendered via built-in string formatting (functionally equivalent, but templates are not externally editable in fallback mode). Jinja2 can be installed via install.py (optional step).

1. Input Analysis

• Task ID: $1
• Status: $2
• Git State: !git diff --staged --stat (summary) AND !git diff --staged (details)

2. Context Dict Construction

Build a context dict matching output_schema.json. The dict MUST contain:

{
    "task_id": "...",
    "status": "...",
    "date": "YYYY-MM-DD",
    "qa_pairs": [
        {"question": "...", "answer": "...", "decision": "..."}
    ],
    "file_modifications": [
        {
            "file_path": "...",
            "summary": "...",
            "reason": "...",
            "role": "...",
            "ripple_effects": ["..."],
            "logic_explanation": "L42-L58: ..."
        }
    ],
    "systemic_impact": {
        "data_flow": "...",
        "functional_hierarchy": "...",
        "framework_impact": "...",
        "api_consistency": "...",
        "performance": "..."
    },
    "verification_status": {
        "tests_passed": ["..."],
        "manual_checks": ["..."]
    }
}

3. Report Generation

Use render.save_changelog(project_root, context) to generate and persist the change log to .claude/temp_log/.

4. Content Standards (Strict)

You MUST adhere to the following 4 rules when writing the log:

1. Completeness (No Token Saving): DO NOT summarize, compress, or omit technical details to save space/tokens. You MUST preserve the FULL technical context.
2. Negative Knowledge (Falsification): You MUST document refuted hypotheses and failed attempts. Explaining "why approach X failed" is as critical as "how approach Y succeeded".
3. Style (Objective): Use formal, simple indicative sentences (Subject + Verb + Object). STRICTLY PROHIBIT unnecessary Adjectives, Adverbs, and Metaphors.
4. Epistemic Humility: DO NOT declare "Fixed" or "Solved" without empirical data (logs/tests). Use "Implemented" or "Attempted" for unverified changes.

5. Execution Rule

• No Adjectives: Use "Implemented X", not "Successfully implemented X".
• No Innocence Presumption: Document risks even if tests passed.

6. Strict Schema Compliance (Implicit)

You MUST read ~/.claude/skills/remy-changelog/output_schema.json to understand the required verification depth. Do NOT output the JSON block. Populate the context dict to satisfy the schema's rigor.