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 rules/stackhawk/agent-skills/stackhawk-hawkscan-scan-qualitygit clone --depth 1 https://github.com/stackhawk/agent-skillsWhat 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 | $0.00076 | $0.03448 |
| Opus 5 | $0.00038 | $0.01724 |
| Sonnet 5 | $0.00015 | $0.00690 |
| Haiku 4.5 | $0.00008 | $0.00345 |
Grade A, and why
stackhawk-hawkscan-scan-quality scanned grade A 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 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
| 1b | Base-path resolve | Whether spec-derived paths actually resolve against the app, or systematically 404 because the config and spec disagree on the base/context path | In the scanned-URI list, look for the same rou How it starts
The opening of the file, as written. The whole thing — 193 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Scan Quality Reference
Run this gate after every scan — full scans and rescans alike — before findings are parsed into fix tasks (SKILL.md calls it Step 4.5, sitting between Step 3's scan run and Step 4's fix-task generation). It is a feedback loop into config tuning, not a governor: gate state never blocks a finding from being reported and fixed. A thin scan that only reached a fraction of the surface can still have found something real, and that finding gets fixed now, regardless of what the gate says. The gate's job is narrower: notice when a scan almost certainly missed API surface, and drive a bounded, additive config fix so the next scan reaches more of it.
The gate derives its expectation fresh every time it runs. There is no persisted plan or prior-run state to compare against — recompute the expected surface from the repo and the effective config each time, the same way discovery would if run again right now.
Multi-config execution. When a repo has more than one stackhawk.yml surface (discovery
writes one config per surface, ordered by DAST value — primary or backing API first, a SPA
frontend last), scan and gate sequentially, one config at a time, in that same
value order. Gate immediately after each config's scan rather than batching every scan
first and gating afterward — an auth-wall caught and fixed on config 1 saves the identical
wasted scan on config 2. The iteration caps described below (2 interactive / 1 autonomous)
apply per config, not to the run as a whole. The autonomous post-code-change loop scans
only the config(s) whose surface the change touched — that scan must still cover that
surface's entire scope (the fix-ALL-findings rule from Step 4 is unchanged) — defaulting to
the primary surface's config when it's unclear which surface a change touched. A full sweep
across every config is an interactive operation only; never trigger one autonomously.
Contents
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 · 193 lines · 76 tokens per session scan A 2b3325c4ddec
stackhawk-hawkscan-scan-quality is a cursor rule published in the GitHub repository stackhawk/agent-skills (16 stars, last pushed 12d ago), licensed MIT. It adds 76 tokens to every session and 3,448 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other cursor rules, from other repositories
angular-20
This rule provides comprehensive best practices and coding standards for Angular development, focusing on modern TypeScript, standalone components, signals, and performance optimizations.
dev-standard
Apache Superset development standards and guidelines for Cursor IDE.
typescript
Changes to these high-fan-out internals can affect every message, delta, element, or rerun. Keep work in them minimal, and benchmark changes with representative stress-test apps.
coolify-ai-docs
Master reference to all Coolify AI documentation in .ai/ directory.
python_lib
Tips and guidelines specific to the development of the Streamlit Python library, not applicable to scripts and e2e tests.
specs
This directory contains product and tech specs for Streamlit features.