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/ackeskin/contexture/new-hookgit clone --depth 1 https://github.com/AcKeskin/contextureWhat 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 | $0.00028 | $0.00316 |
| Opus 5 | $0.00014 | $0.00158 |
| Sonnet 5 | $0.00006 | $0.00063 |
| Haiku 4.5 | $0.00003 | $0.00032 |
Grade B, and why
new-hook scanned grade B with 1 finding 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 yesterday.
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.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
The skill picks a recipe (one of `bash-command-blocker`, `file-write-blocker`, `mcp-tool-blocker`, `context-injector`, `rule-prime`, `session-recovery-advisory`), prompts for a hook name, asks one question per recipe pla What it actually says
Run the new-hook skill for the current context.
Any text after /new-hook is treated as a hint that biases recipe selection but never enough to skip the interview. Examples:
/new-hook— full interview from the recipe-selection step./new-hook block writes to secrets/— hint suggestsfile-write-blocker; skill still confirms./new-hook PreToolUse on Bash— hint suggestsbash-command-blocker; skill still confirms.
The skill picks a recipe (one of bash-command-blocker, file-write-blocker, mcp-tool-blocker, context-injector, rule-prime, session-recovery-advisory), prompts for a hook name, asks one question per recipe placeholder, detects whether to write into contexture/hooks/ or ~/.claude/hooks/ based on bootstrap state, shows a settings.json diff for confirmation, and only after y writes the four artefacts (hook + block fixture + allow fixture + Node-based runner) and merges the registration into ~/.claude/settings.json.
After writing, the skill spawns the runner and reports PASS/FAIL for both fixtures. Fail-noisy: artefacts stay on disk; the skill does not auto-revert.
See ~/.claude/skills/new-hook/SKILL.md for the full procedure.
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.
- yesterday First seen · 18 lines · 28 tokens per session scan B f50fb4a5f07b
new-hook is a command published in the GitHub repository AcKeskin/contexture (2 stars, last pushed 1mo ago), licensed MIT. It adds 28 tokens to every session and 316 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
translate-review-to-single-human
Translate a multi-reviewer code review into a single human-voice GitHub PR comment.
address
Address code review feedback — corroborate, validate, and implement changes from a review's final.md.
map
Generate a Code Review Map to help navigate large, complex changesets.
post
Post the current OCR review to a GitHub PR.
show
Display a past OCR review session.
history
List past OCR review sessions.