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 UnboundCompute/security-agent-skills --skill hunting-bugs-with-a-code-graphgit clone --depth 1 https://github.com/UnboundCompute/security-agent-skillsWrote 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/unboundcompute/security-agent-skills/hunting-bugs-with-a-code-graph)<a href="https://agentmods.dev/skills/unboundcompute/security-agent-skills/hunting-bugs-with-a-code-graph"><img src="https://agentmods.dev/badge/skills/unboundcompute/security-agent-skills/hunting-bugs-with-a-code-graph/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/unboundcompute/security-agent-skills/hunting-bugs-with-a-code-graph"><img src="https://agentmods.dev/badge/skills/unboundcompute/security-agent-skills/hunting-bugs-with-a-code-graph.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 | $0.00127 | $0.01468 |
| Opus 5.5 | $0.00051 | $0.00587 |
| Sonnet 5.5 | $0.00025 | $0.00294 |
| Haiku 4.5 | $0.00013 | $0.00147 |
Grade A, and why
hunting-bugs-with-a-code-graph 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 1mo 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 — 119 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Hunting bugs with a code graph
Grep finds strings; it misses the caller behind a rename, an alias, or an import
indirection, and it cannot answer "what can flow into this argument." Reasoning
over a codebase's structure - its call graph and dataflow - can. This skill is
the master loop for a source-level hunt. Two companions go deeper on single
moves: adjudicating-taint-paths (lead → decided finding) and
auditing-guard-gaps (the unguarded peer of a guarded function).
When to use
- You have source and want coverage, not a one-off keyword search.
- You're cold on an unfamiliar codebase and need to find where input arrives.
- You want to work a bug taxonomy systematically and prove what you ruled out.
Scope check (do this first)
Only run on code you're authorized to analyze: your own, an OSS project you contribute to, a CTF, or an engagement where source review is in scope. If you can't name why you're allowed to read this source, stop.
The loop
-
Index the target. Build a structural index of the source tree. It is a snapshot - re-index whenever the code changes in a way that matters, or you will adjudicate against stale structure.
-
Orient before hunting. Start from the most-connected functions - the structural spine, where input arrives and trust boundaries sit - not from a file you happened to open. Map the top-level entry points and the module layout before drilling in.
-
Enumerate the whole taxonomy - never one family. List every bug class your catalog covers before examining any single one. Do not scope the hunt to the class you expect (injection, memory copy, whatever) because it's familiar. Coverage is over the whole taxonomy; a hunt that only ever looks at one family is not a hunt, it's a confirmation of a hunch.
-
Rank is triage, not a filter. A ranked lead list orders your attention. Inclusion is exhaustive; a low rank never justifies dropping a candidate from examination. Work down the list - don't truncate it.
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.
- 1mo ago First seen · 119 lines · 127 tokens per session scan A a0e429ec3157
hunting-bugs-with-a-code-graph is a skill published in the GitHub repository UnboundCompute/security-agent-skills (5 stars, last pushed 1mo ago), licensed MIT. It adds 127 tokens to every session and 1,468 once invoked, about $0.0005 per session on Opus 5.5. 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 2026-09-05.
Other skills, from other repositories
ffuf-claude-skill
Skill "ffuf-claude-skill" from Agent-Threat-Rule/agent-threat-rules, covering ffuf claude skill, when to use this skill and instructions.
codeinspectus-fix-one
Investigate and remediate exactly one user-selected CodeInspectus finding with evidence-gated reproduction, a separately approved minimal patch, focused regression testing, and an exact-prior-scan rescan. Use when a user asks an agent to examine, reproduce, fix, or verify one CodeInspectus finding without batching…
gauntlex:doctor
Full GAUNTLEX environment health check — model reachable, ChromaDB writable, AVF fixtures pass, no unexpected outbound network (air-gap verification). Use before first run or when debugging failures.
report-error
Verify a suspected plugin defect locally, prepare a minimal anonymised GitHub issue, and publish only after the user reviews and explicitly approves its exact content. Also supports a local-only diagnostic bundle.
prodcheck-review
Review this codebase against the prodcheck pre-production checklists — security, performance, scale, integrations, post-launch readiness and people. Use when asked to check whether a project is ready to ship, to audit an area before launch, or to work through a specific checklist. Produces evidence with file:line…
diagnose-bundle
Maintainer/dev skill that triages an anonymised diagnostic bundle (appsec-diag-.tgz, produced by scripts/runtime/diagnosticbundle.py) a user sent after a pipeline failure. Runs the deterministic inspect, then cross-references the plugin source (scripts/, compact runtimes, agents/, AGENTS.md) and known-bug history to…