القائمة
Use when maintaining, validating, or checking boundaries of the Mathodology skills repository itself, or deciding whether a change belongs on this branch.
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
استنادا إلى تصنيف SOC المهني
Use when running, backing up, transferring, restoring, archiving, or globally installing the Mathodology skills-only repository and its award workflows.
Use when maintaining archived knowledge about Mathodology's former Python agent pipeline, worker roles, prompts, Coder execution, HMML, MATLAB, search, or critic behavior.
Use when running Mathodology award-workflow phase gates, judge panels, structured handoffs, figure QA, or rendered-PDF QA in a contest run.
Use when validating the Mathodology skills-only repository or preserving archived knowledge about the former development, testing, deployment, packaging, and release workflows.
Use when adding, updating, validating, or reviewing Mathodology project skills, SKILL.md files, or agents/openai.yaml metadata.
Use when maintaining archived knowledge about Mathodology's former gateway, API routes, auth, state, LLM routing, exports, submission bundles, or contracts.
| name | mathodology-project-orientation |
| description | Use when maintaining, validating, or checking boundaries of the Mathodology skills repository itself, or deciding whether a change belongs on this branch. |
This branch is a skills-only GitHub tree. Treat it as an AI-coding knowledge pack, not as the original runnable Mathodology application.
Current work should normally edit only:
.claude/skills/<skill-name>/ — its SKILL.md,
agents/openai.yaml, and a scripts/ directory when the skill ships one
(e.g. .claude/skills/mathodology-award-gates/** and
.claude/skills/mathodology-dev-test-release/scripts/validate_repo.py).claude/agents/<agent-name>.md (including .claude/agents/mathodology-award-judge.md).claude/workflows/<workflow-name>.md.claude/skills/mathodology-whole-project/scripts/create-source-backup.shAGENTS.mdREADME.mdREADME_en.mddocs/SKILLS.mddocs/SKILLS_zh.mddocs/INSTALL.mddocs/INSTALL_zh.mddocs/WORKFLOWS.mddocs/WORKFLOWS_zh.mddocs/BACKUP.md.gitignoreLICENSE is retained but should not change unless the license changes.
The current GitHub tree should not contain application source, generated clients, CI, deployment, package-manager, installer, data, or test trees.
If a task requires historical implementation detail, use Git history or another branch in a separate worktree. Do not add those files back to this branch as part of ordinary skills maintenance.
Before publishing cleanup or skill maintenance, verify tracked files stay inside the skills-repository whitelist:
python3 .claude/skills/mathodology-dev-test-release/scripts/validate_repo.py whitelist
Run validate_repo.py all for the full set of maintenance gates (skills,
metadata, links, whitelist, agents, sync). The whitelist logic lives only in
that script; do not re-inline it here.
Search for stale current-path claims before committing (requires ripgrep; the
grep -E fallback is shown second):
rg -n "apps/|crates/|packages/|docs/skills|Dockerfile|docker-compose|just |cargo|pnpm|uv run|\.github|installer" README.md README_en.md AGENTS.md docs .claude/skills
grep -REn "apps/|crates/|packages/|docs/skills|Dockerfile|docker-compose|just |cargo|pnpm|uv run|\.github|installer" README.md README_en.md AGENTS.md docs .claude/skills
Hits are acceptable only when they describe removed historical material or an absence check. They must not instruct agents to edit or run missing current files. The current skills, scripts, and award-gate content do not match this pattern, so any hit is a stale claim to review.
mathodology-whole-project.mathodology-skill-authoring.docs/WORKFLOWS.md.mathodology-award-gates.mathodology-dev-test-release.