audit-augmentation — Claude Code skill: what it does, how to install
Projects SARIF static analysis reports, weAudit notes, and binary graphs onto Trailmark code graphs. Useful when mapping vulnerabilities to call graphs and data taint paths.
npx skills add trailofbits/skills --skill audit-augmentation -a claude-codeWhat audit-augmentation does
The skill maps external audit findings onto Trailmark code graphs. It matches findings from SARIF files generated by Semgrep or CodeQL alongside weAudit annotations by checking file paths and line ranges. The engine then builds severity subgraphs such as sarif:error or weaudit:high.
It does not run static analysis tools or construct the base graph. You must run engine.preanalysis() before loading findings. Skipping pre-analysis breaks cross-referencing with taint propagation and blast radius. That defeats the purpose of graph augmentation.
You need the Trailmark tool installed via uv. Binary graph ingestion requires Trailmark 0.4.0 or newer through engine.augment_binary(). Starting with version 0.5.0, links to external binary symbols can be declared inside .trailmark/links.toml.
What it helps with
- project Semgrep and CodeQL SARIF results onto a code graph
- import manual auditor notes from weAudit files
- isolate functions with high-severity issues using severity subgraphs
- cross-reference static findings with blast radius and taint reachability
- load external binary graph exports into the source code graph
Example prompts for the agent
- Import results.sarif into the Trailmark graph and show functions with high-severity findings
- Overlay audit notes from .vscode/alice.weaudit onto the graph and check overlap with tainted nodes
- Run trailmark augment on the target directory with results.sarif and output JSON
How to install audit-augmentation
1Skills CLI (Vercel)
npx skills add trailofbits/skills --skill audit-augmentation -a claude-code2Manually: put the files in place
~/.claude/skills/audit-augmentation/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/audit-augmentation/
1Skills CLI (Vercel)
npx skills add trailofbits/skills --skill audit-augmentation -a codex2Manually: put the files in place
.agents/skills/audit-augmentation/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add trailofbits/skills --skill audit-augmentation -a cursor2Manually: put the files in place
.cursor/skills/audit-augmentation/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install trailofbits/skills audit-augmentation2Manually: put the files in place
.github/skills/audit-augmentation/SKILL.md1Skills CLI (Vercel)
npx skills add trailofbits/skills --skill audit-augmentation -a gemini-cli2Platform CLI
gemini skills install https://github.com/trailofbits/skills --path plugins/trailmark/skills/audit-augmentationManually: put the files in place
.devin/skills/audit-augmentation/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add trailofbits/skills --skill audit-augmentation -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
From the same set35
The whole set →Similar
Questions about audit-augmentation
How to install audit-augmentation in Claude Code?
Run npx skills add trailofbits/skills --skill audit-augmentation -a claude-code. If Trailmark is missing, install it with uv tool install trailmark.
What does audit-augmentation do?
It maps SARIF analysis outputs, weAudit notes, and binary graph exports onto Trailmark code graphs. This groups issues into severity subgraphs and lets you evaluate their impact across connected functions.
Why must preanalysis run before importing findings?
Without pre-analysis, the graph lacks blast radius and taint reachability data. Running engine.preanalysis() lets you cross-reference detected vulnerabilities with data-flow paths.