Trigger Doctor — Testing skill for Claude Code
🩺 Behavioral testing for skill triggers — diagnose why a skill never fires (or fires too often), prescribe a fixed description, save a regression suite.
How to install Trigger Doctor
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open mohamed-khairy-5i/trigger-doctor and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Trigger Doctor does
🩺 Behavioral testing for skill triggers — diagnose why a skill never fires (or fires too often), prescribe a fixed description, save a regression suite. One SKILL.md, 79 agents (Claude Code, Hermes, Codex, Gemini...)
Alternatives in Testing
- Superpowers Evals — Behavioral eval lab (Quorum) for the superpowers project that drives real coding-agent CLIs (Claude, Codex, Ge 107 ★
- Skill Trigger Test — PHẢI dùng khi: vừa tạo agent/skill mới, vừa sửa description frontmatter, nghi ngờ routing conflict giữa 2 agen 90 ★
- Flaky Test — Diagnose a flaky test by running it N times 68 ★
README
trigger-doctor
🩺 Diagnose your agent skills — behavioral testing, prescriptions and regression suites
**Behavioral testing for skill triggers.** Most skills don't die of broken logic — they die of never being opened. The `description` field is the gate agents use to decide whether to load a skill, and almost nobody tests it.
trigger-doctor is an open-source [Agent Skills](https://agentskills.io) skill that diagnoses whether another skill will actually trigger, finds *why* it doesn't, prescribes a fixed description, and saves a regression suite you re-run after every model update. It works with any runtime that follows the open standard — **one `SKILL.md`, 79 agents** (Claude Code, Hermes, Codex, Cursor, Gemini CLI, GitHub Copilot, …).
Quick start
npx skills add mohamed-khairy-5i/trigger-doctor
One command installs it for **all 79 agents** — the CLI knows each agent's skills directory. Then ask your agent: *"check my skill"*.
**Where it lands (global):**
| Agent | Directory |
|---|---|
| Cla |
Related Skills
Tester
Spec-driven test generation and cross-language behavioral verification for apcore ecosystem — ONLY when user e
Draft Filter
Draft, test, and optionally save a FeedSheriff filter from a plain-English description.
Canary Debug Flake
Diagnose a flaky test and propose a deterministic fix from code, CI logs, or a failure description.
Continuity Test
Memory retrieval regression test — runs curated query→memory cases through hybrid search (use: /continuity-tes
Test Judge
Test whether an AI judge agrees with you, then estimate how often the product really fails
Gtm QA
Run autonomous QA verification on any GTM deployment — pixel fires, funnel walkthroughs, mobile testing, perfo
Related Agents
Arch Evaluator
Runs the eval harness against a newly generated plugin — skill trigger accuracy, agent delegation correctness,
Dead Code Sweep
Periodic sweep, within one Zanix repo at a time, for orphaned modules (no longer on the real path but still pr
Doctor
Diagnose AgentLint configuration and hook installation issues