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 agentmods add commands/cafitac/hermit-agent/code-apply-hermitgit clone --depth 1 https://github.com/cafitac/hermit-agentWhat 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 | $0.00000 | $0.01173 |
| Opus 5 | $0.00000 | $0.00587 |
| Sonnet 5 | $0.00000 | $0.00235 |
| Haiku 4.5 | $0.00000 | $0.00117 |
Grade A, and why
code-apply-hermit 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 2d 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 — 158 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/code-apply-hermit — Claude reads the review, Hermit applies it
Hands the P1–P5 findings from the most recent /code-review to Hermit
so Claude doesn't burn tokens re-reading every file and retyping every
edit. Claude stays on judgment; Hermit does the mechanical apply.
Arguments
$ARGUMENTS — PR number, optionally followed by severity filters.
Examples: 123, 123 P1 P2 P3.
--model <model_name>— which executor model Hermit should run. Defaults to the gateway's active model (./bin/gateway.sh --statusto see the current pick).
Prerequisites
- A
/code-reviewoutput from the previous turn in this conversation. - Hermit gateway + MCP server running.
Workflow
Step 1 — locate the worktree
If the argument contains a PR number, follow the repo's worktree convention to resolve the worktree path (or infer from the current branch if you are already inside it). If no worktree / PR info can be resolved, ask the user which directory to operate on.
Step 2 — harvest review findings
Scan the prior conversation for the last /code-review output and
extract each finding: severity, file path, line number, description,
and the proposed direction.
If no review output is present, stop and tell the user to run
/code-review $ARGUMENTS first.
Step 3 — choose what to apply
- Default: every P1 through P5 item.
- Filter: if severities are listed in
$ARGUMENTS(e.g.P1 P2), apply only those.
Step 3.5 — write findings to disk
Save the selected findings to:
.omc/reviews/<branch>-review-<YYYYMMDDHHMMSS>.md
Format:
# Code Review Findings — <branch>
## P1
- `path/to/file.py:42` — short description — suggested fix
## P2
- ...
Step 4 — concurrency lock
mkdir -p .hermit
cat > .hermit/active-task.lock <<EOF
{
"task_id": "<filled-after-run_task>",
"started_at": "$(date -Iseconds)",
"skill": "code-apply-hermit",
"pr": "$ARGUMENTS",
"cwd": "<worktree_path>",
"findings": "<findings_file_path>"
}
EOF
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.
- 2d ago First seen · 158 lines · 0 tokens per session scan A ee66af184cab
code-apply-hermit is a command published in the GitHub repository cafitac/hermit-agent (4 stars, last pushed 6d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,173 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 commands, from other repositories
toefl-grade
TOEFL 답안/녹음 채점 — 로컬 LLM(Ollama)과 whisper.cpp로 4개 영역을 채점하여 SCORES.md에 점수를 누적 기록한다.
toefl-practice
TOEFL 유형 문제 생성 — 로컬 LLM(Ollama)이 reading/listening/speaking/writing 영역의 실전형 문제를 생성하여 practice/에 저장한다.
toefl-drill
취약 영역 집중 반복 — SCORES.md에서 약점을 식별하여 해당 유형의 문제를 연속 생성/채점하는 딥 드릴 루프를 안내한다.
toefl-roadmap
이번 주 토플 학습 로드맵/과제 표시 — schedule.yaml에서 사용자가 지정한 학습 기간·요일·시간대·주차별 목표를 읽어 현재 날짜 기준으로 안내한다.
toefl-status
토플 학습 현황 요약 — schedule.yaml(D-day/목표점수) + JOURNAL.md/SCORES.md를 파싱하여 진행률, 영역별 점수 추이, 목표 대비 갭, 시험일까지 남은 일수를 표시한다.
opencode
Manage OpenCode/Zen connections, switch models, and configure local Ollama.