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 forefy/.context --skill sandboxed-audit-runnergit clone --depth 1 https://github.com/forefy/.contextWrote 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/forefy/.context/sandboxed-audit-runner)<a href="https://agentmods.dev/skills/forefy/.context/sandboxed-audit-runner"><img src="https://agentmods.dev/badge/skills/forefy/.context/sandboxed-audit-runner/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/forefy/.context/sandboxed-audit-runner"><img src="https://agentmods.dev/badge/skills/forefy/.context/sandboxed-audit-runner.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high Privilege Escalation · line 52 Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.Fix: Remove references to credential paths. Use environment variables or secrets managers. For docs, use placeholder paths (e.g., /path/to/config). Never load .env or token files in production code paths.
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 | $0.00037 | $0.00599 |
| Opus 5 | $0.00018 | $0.00300 |
| Sonnet 5 | $0.00007 | $0.00120 |
| Haiku 4.5 | $0.00004 | $0.00060 |
Grade B, and why
sandboxed-audit-runner scanned grade B with 2 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Reaches for credential filesmediumPrivilege escalation
SSH keys, cloud credentials, git-credentials, .npmrc, /etc/shadow: reading these is how a config file becomes a credential leak.
cat ~/.ssh/id_rsa # should be blocked Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl https://example.com # should be blocked if not in allowedDomains How it starts
The opening of the file, as written. The whole thing — 72 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Sandboxed Audit Runner
Before running any audit on untrusted code, this skill configures and launches srt to wrap the entire agent session. The goal is to protect the host from prompt injection attacks embedded in the codebase under review - malicious comments, filenames, or configs designed to make the auditing agent exfiltrate SSH keys, tokens, or make unauthorized network calls.
SKILL DIRECTORY DETECTION:
SKILL_DIR=$([ -d "$HOME/.context/skills/sandboxed-audit-runner" ] && echo "$HOME/.context/skills/sandboxed-audit-runner" || echo ".context/skills/sandboxed-audit-runner")
Prerequisites
npm install -g @anthropic-ai/sandbox-runtime
srt --version
On Linux, also install bubblewrap: apt install bubblewrap.
Setup
1. Create the audit session config
Read $SKILL_DIR/references/smart-contract.srt.md, extract the JSON block, and write it to .srt-audit.json in the project root. Remove any allowedDomains entries the audit does not need.
2. Start the sandboxed session
srt --settings .srt-audit.json bash
All commands run inside that shell - including any invoked audit skill - are now sandboxed at the OS level. On macOS this uses sandbox-exec; on Linux it uses bubblewrap.
3. Invoke the audit skill normally
With the sandbox active, invoke the audit skill as usual:
run smart-contract-audit
The entire agent conversation and all tool calls it makes will execute under the srt restrictions for the duration of the session.
4. Verify the sandbox is active (optional)
cat ~/.ssh/id_rsa # should be blocked
curl https://example.com # should be blocked if not in allowedDomains
5. Clean up
exit
rm .srt-audit.json
Why this matters
Audited code can contain:
- Comments or string literals with prompt injection payloads targeting the AI
package.jsonpostinstall scripts that probe~/.awsor~/.ssh- Test fixtures or build scripts with embedded exfiltration logic
- Configs that instruct tooling to phone home
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 6d ago Changed · +1 lines 65a1d57fa0ab
- 10d ago First seen · 71 lines · 37 tokens per session scan B f026f48efa1a
sandboxed-audit-runner is a skill published in the GitHub repository forefy/.context (144 stars, last pushed today), licensed MIT. It adds 37 tokens to every session and 599 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it B with 2 findings (reaches for credential files, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
eresus-manual-security-audit
Elite manual security code review skill for deep, adversarial vulnerability hunting and exploit-chain discovery. Trigger when the user asks to: "do a deep security audit", "manual code review", "find exploit chains", "hunt for logic bugs", "red-team this codebase", "do an offensive security review", "review this like…
eresus-sast-scanner
General-purpose Static Application Security Testing (SAST) skill for code vulnerability analysis. Trigger when the user asks to: "analyze code for vulnerabilities", "review code security", "find security bugs", "do a SAST scan", "check for [vulnerability type] in code", "audit source code", or requests a security code…
eresus-php-audit
Deep PHP-specific security audit skill covering injection, deserialization, file operations, auth bypass, POP chain discovery, and CMS-specific patterns. Trigger when auditing PHP code: "audit this PHP app", "find PHP security issues", "check Laravel/WordPress for vulnerabilities", "PHP SAST review", "check for PHP…
eresus-python-audit
Deep Python-specific security audit skill with 50+ vulnerability class coverage across 7 categories. Trigger when auditing Python code: "audit this Python app", "find Python security issues", "check Flask/Django for vulnerabilities", "Python SAST review", "check for pickle vulnerabilities", "review this FastAPI code".…
eresus-remediator
Security remediation skill for fixing confirmed or likely SAST findings in source code. Trigger when the user asks to: "fix a vulnerability", "patch this security bug", "remediate SAST findings", "harden this endpoint", "make this auth flow safe", or wants code changes that remove a confirmed security issue while…
eresus-variant-analysis
GHSA/CVE variant analysis workflow for finding similar vulnerability patterns across a codebase. Trigger when the user asks to: "find variants of this CVE", "GHSA variant analysis", "find similar bugs", "hunt for the same pattern", "are there other places with this vulnerability?", or when a known vulnerability is…