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.
npx skills add ForgeyClap/claude-forge --skill forge-debuggit clone --depth 1 https://github.com/ForgeyClap/claude-forgeWrote 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/skills/forgeyclap/claude-forge/forge-debug)<a href="https://agentmods.dev/skills/forgeyclap/claude-forge/forge-debug"><img src="https://agentmods.dev/badge/skills/forgeyclap/claude-forge/forge-debug/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/skills/forgeyclap/claude-forge/forge-debug"><img src="https://agentmods.dev/badge/skills/forgeyclap/claude-forge/forge-debug.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 | $0.00000 | $0.01118 |
| Opus 5.5 | $0.00000 | $0.00447 |
| Sonnet 5.5 | $0.00000 | $0.00224 |
| Haiku 4.5 | $0.00000 | $0.00112 |
Grade A, and why
forge-debug 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 20d 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 — 63 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Forge playbook — Systematic debugging
Do not duplicate ECC skills — defer to: systematic-debugging (the deep methodology reference; it ships with Forge as a pinned vendored skill). This
file is the Forge-specific orchestration wrapper: it binds that methodology to the honesty core and to
which Boss does what.
Hard rules
- Classify the error first. Build-blocking, type, configuration, or dependency — fix build-blocking errors first before moving to other issues.
- The Iron Law: no fix without root-cause investigation first. A patch that only makes the symptom go away, without an explanation of why the bug happened, is not a fix — it is a guess.
- Never claim "fixed" without proof. Per the project honesty core (
CLAUDE.md→ Honesty), a bug is only "fixed" when there is a real regression test that failed before the change (RED) and passes after it (GREEN). A verbal claim of "should be fixed now" is not evidence. - Minimal-diff check after each fix. After each fix, re-run the same diagnostic and confirm only the targeted error changed.
- Do not silently patch around a failure (swallowed exception, widened try/catch, disabled test, loosened assertion) to make a check go green — that is a rework-flag, not a fix.
- If root cause can't be found within a reasonable number of attempts, say so honestly (
BLOCKED/needs owner input) rather than shipping a guess dressed up as a fix.
The loop (reproduce → isolate → root-cause → fix → regression test → verify)
- Reproduce. Get a minimal, reliable repro (exact command, exact input, exact error). If it can't be reproduced, that itself is a finding — don't invent a story around a bug you haven't actually seen.
- Isolate (bisect / binary-search). Narrow the surface: which commit introduced it (
git bisect), which input triggers it (binary-search the input space), which layer owns it (add targeted logging / breakpoints, remove code paths until the symptom disappears). Prefer removing variables over adding theories. - Root-cause, not symptom. State the mechanism in one sentence: "X happens because Y does Z under condition W." If you can't state it that precisely, you don't have the root cause yet — keep isolating.
- Fix at the root, matching the existing architecture and conventions (follow the project's own conventions — a
coding-style.mdor CLAUDE.md rules when the project has them) — not a parallel workaround bolted on beside the real cause. - Regression test. Write a test that fails on the pre-fix code and passes on the post-fix code. It stays in the suite permanently — it is the proof, and it prevents the same bug returning silently.
- Verify. Re-run the full relevant suite (not just the new test) to confirm no other case regressed; quote the real command and its real output.
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.
- 20d ago Changed · +3 lines · -42 tokens per session 0c9247cdfd3f
- 26d ago Changed 27848ef992ed
- 28d ago Changed cde8f2ccf615
- 1mo ago First seen · 60 lines · 42 tokens per session scan A 8571c36e6f9d
forge-debug is a skill published in the GitHub repository ForgeyClap/claude-forge (2 stars, last pushed 21d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,118 tokens. 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 2026-08-31.
Other skills, from other repositories
systematic-debugging
4-phase root cause debugging: understand bugs before fixing.
module-regression
Govern cross-module regression in large repositories with a REGRESSION.md ledger that records each module's downstream consumers and executable acceptance command. After a change, run the changed module and every affected downstream command, and use observed exit codes as regression evidence under the project's…
langsmith-observability
LLM observability platform for tracing, evaluation, and monitoring. Use when debugging LLM applications, evaluating model outputs against datasets, monitoring production systems, or building systematic testing pipelines for AI applications.
test-failure-explorer
Import and query an existing Vitest 3.2 JSON report to explore test failures, file-level errors and original diagnostic evidence.
test-flakiness
Find flaky tests from CI logs — aggregates pass rates, spots intermittent failures, recommends quarantine. After multiple runs.
maintain-formal-verification-skill
Bring a project's formal verification back in line with the code after the code changes: find the commits that touched verified code, update the model, harness, or properties to describe the code as it is now, rerun, and say whether each failure is drift in the verification or a bug in the code. Use for…