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/Intense-Visions/harness-engineeringWrote 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/rules/intense-visions/harness-engineering/respect-hooks)<a href="https://agentmods.dev/rules/intense-visions/harness-engineering/respect-hooks"><img src="https://agentmods.dev/badge/rules/intense-visions/harness-engineering/respect-hooks.svg" alt="Measured on agentmods" 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 | $0.00514 | $0.00514 |
| Opus 5 | $0.00257 | $0.00257 |
| Sonnet 5 | $0.00103 | $0.00103 |
| Haiku 4.5 | $0.00051 | $0.00051 |
Grade A, and why
respect-hooks 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 3d 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 — 38 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Respect lifecycle hooks
The standard hook profile (shipped with the harness plugin) wires several pre- and post-tool hooks that exist for specific safety reasons:
| Hook | Event | What it blocks |
|---|---|---|
block-no-verify |
PreToolUse:Bash | Any command containing --no-verify (or equivalent skip flags) |
protect-config |
PreToolUse:Write/Edit | Edits to harness/git/CI config without explicit human approval |
quality-gate |
PostToolUse:Edit/Write | Files left in failing-validation state |
pre-compact-state |
PreCompact | Loss of session state during context compaction |
adoption-tracker / telemetry-reporter |
Stop | (log-only — never block) |
These hooks are not noise. Each was added in response to a specific failure mode.
What you must not do
- Never propose
git commit --no-verify.block-no-verifywill reject it, and proposing it anyway tells the human you didn't read the rule. - Never propose modifying or removing hook scripts without the human asking explicitly. If a hook is firing too often, the right fix is calibration (in the hook's own rules file or
harness.config.json), not deletion. - Never bypass
protect-configby suggesting a manual override. Ifharness.config.jsonor.claude/settings.jsonlegitimately needs editing, do so directly via the Edit tool — the hook will surface for approval, which is the intended flow. - Never silence
quality-gateafter a failed edit. Fix the file so validation passes, or revert the edit.
When a hook misfires
If you genuinely believe a hook is wrong (false positive on a rule):
- Surface it to the human. Don't quietly route around it.
- Suggest updating the hook's logic or calibration, not bypassing the event.
- Reference the specific hook by name so the human can locate and audit it.
Why this rule exists
Hooks are mechanical safety. If a hook is wrong, fix it deliberately. If a hook is right, respect it. The middle path — bypassing a hook because it's inconvenient — is how the original incident that motivated the hook recurs.
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.
- 3d ago First seen · 38 lines · 514 tokens per session scan A a15bc77aacef
respect-hooks is a cursor rule published in the GitHub repository Intense-Visions/harness-engineering (20 stars, last pushed yesterday), licensed MIT. It adds 514 tokens to every session, about $0.0026 per session on Opus 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 cursor rules, from other repositories
git-commit-attribution
Git commits must use the human developer identity only; never Cursor Agent co-authorship.
skill-management
🔒 LOCKED — managed by clade Source: rules/core/skill-management.md Edit at: $CLADEHOME Local edits will be reverted by the next sync. -->.
iris
GitHub operations specialist — branches, pull requests, issues, releases, tags. Called by zeus after review. Never pushes or merges without explicit human approval. Integrates with VS Code GitHub Pull Requests extension.
rebase-only-merged-agents
When rebasing after a merge, only rebase the agents whose branches you just merged. Do not rebase agents that are still running.
git-workflow
Rules for git operations, commit strategy, and GitHub CLI usage.
release-checklist
What to update with each code change or version release.