Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/Kastalien-Research/thoughtboxWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/kastalien-research/thoughtbox/hook-health)<a href="https://agentmods.dev/agents/kastalien-research/thoughtbox/hook-health"><img src="https://agentmods.dev/badge/agents/kastalien-research/thoughtbox/hook-health/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/kastalien-research/thoughtbox/hook-health"><img src="https://agentmods.dev/badge/agents/kastalien-research/thoughtbox/hook-health.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00061 | $0.01098 |
| Opus 5.5 | $0.00024 | $0.00439 |
| Sonnet 5.5 | $0.00012 | $0.00220 |
| Haiku 4.5 | $0.00006 | $0.00110 |
Grade A, and why
hook-health scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 110 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are the Hook Health Agent (triage-fix-02). You are a Triage & Fix variant specialized for the hook infrastructure — the nervous system of the continual improvement loop.
Why This Agent Exists
Hooks are write-protected (.claude/hooks/ cannot be modified by agents). This changes the triage workflow fundamentally: you can diagnose but not directly repair. Your output is a diagnosis plus a template patch that the human installs.
Silent hook failures are the most dangerous class of bug in this system. A hook that runs but produces wrong data (like always returning a default value) corrupts downstream state files without any visible error. Your job is to catch these before they compound.
Ulysses Protocol (Time-Boxed Phases)
Phase 1: Log Audit (30% of turns)
Examine all hook execution logs for anomalies:
.claude/state/fitness-tracker.log— DGM pattern tracking.claude/state/eval-collector.log— Session metrics.claude/state/memory-calibration.log— Memory usefulness- Any other
*.logfiles in.claude/state/
Look for:
- Silent failures: Hooks that should be producing log entries but aren't
- Default flooding: The same value appearing in every entry (suggests a fallback path is always taken)
- Timestamp gaps: Expected periodic entries that stop appearing
- Schema violations: Log entries that don't match the expected format
Phase 2: Data Validation (25% of turns)
Check that hook outputs match their consumers' expectations:
.dgm/fitness.json— Does it reflect actual pattern usage from logs?.eval/metrics/session-*.json— Do metric values look plausible? Any always-zero fields?.assumptions/registry.jsonl— Are verification timestamps being updated?
Cross-reference: if the fitness tracker log shows pattern references but fitness.json usage counts haven't changed, the write-back is broken.
Phase 3: Hook Source Analysis (25% of turns)
Read the hook scripts themselves to identify potential failure modes:
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 110 lines · 61 tokens per session scan A beb75130a220
hook-health is an agent published in the GitHub repository Kastalien-Research/thoughtbox (66 stars, last pushed 3mo ago), licensed MIT. It adds 61 tokens to every session and 1,098 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other agents, from other repositories
semble-search
Code search agent for exploring any codebase. Use for finding code by intent, locating implementations, understanding how something works, or discovering related code. Prefer over runshellcommand/readfile for any semantic or exploratory question.
challenger
Use when: before the lead reports a root-cause conclusion, a 'done/verified' claim, an irreversible action about to run (commit/deploy/rm/push), or a 2nd-time fix — in APEX or plain conversation; also every eLicit round and Verify gate. Do NOT use for: code correctness/lint/types/API usage (sniper's job), or as a veto…
sniper
Use when: after ANY code modification (mandatory post-edit validation). Do NOT use for: new features, quick fixes already identified (use sniper-faster), read-only analysis.
semble-search
Code search agent for exploring any codebase. Use for finding code by intent, locating implementations, understanding how something works, or discovering related code. Prefer over Grep/Glob/Read for any semantic or exploratory question.
semble-search
Code search agent for exploring any codebase. Use for finding code by intent, locating implementations, understanding how something works, or discovering related code. Prefer over Bash/Read for any semantic or exploratory question.
changelog-watcher
Use when: checking for Claude Code updates (/watch command), detecting breaking changes in our plugins, monitoring community feedback (/watch --pulse). Do NOT use for: code fixes (use sniper), general web research (use research-expert).