VulnHunter is an AI-assisted source-code security tool that examines software like an attacker to determine which defects can actually be exploited. It is for finding vulnerabilities, tracing possible attack paths, and suggesting fixes, especially when used with Claude Code and Claude Opus.
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 capitalone/VulnHunter --skill vulnhunter-fixgit clone --depth 1 https://github.com/capitalone/VulnHunterWrote 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/capitalone/vulnhunter/vulnhunter-fix)<a href="https://agentmods.dev/skills/capitalone/vulnhunter/vulnhunter-fix"><img src="https://agentmods.dev/badge/skills/capitalone/vulnhunter/vulnhunter-fix/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/capitalone/vulnhunter/vulnhunter-fix"><img src="https://agentmods.dev/badge/skills/capitalone/vulnhunter/vulnhunter-fix.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 4 findings, 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 223 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.
- medium Data Exfiltration · line 168 Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
- medium Data Exfiltration · line 175 Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
- medium Excessive Agency · line 346 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
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 | $0.00135 | $0.07866 |
| Opus 5.5 | $0.00054 | $0.03146 |
| Sonnet 5.5 | $0.00027 | $0.01573 |
| Haiku 4.5 | $0.00014 | $0.00787 |
Grade A, and why
vulnhunter-fix 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 2mo 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.
- `gh issue list` fails → trying `git ls-remote` or `curl https://api.github.com/...`. How it starts
The opening of the file, as written. The whole thing — 407 lines — stays where its author put it; the contents beside it link to each section on GitHub.
VulnFix — Automated Security Remediation via TDD
Overview
VulnFix takes VulnHunter scan results and automates the full remediation lifecycle:
- Parse findings from the report (or from
vulnhunter-labeled issues in in-place mode) - Plan fix order and approach
- For each vulnerability: write exploit demo → write failing test → implement fix → verify test passes
- Deliver as PRs — to a private fork (fork mode) or to the same repo (in-place mode)
Modes
The skill runs in one of two modes, dispatched automatically.
| Mode | Trigger | Workflow |
|---|---|---|
| In-place (default when invoked from inside a target repo checkout) | User runs /vulnhunter-fix with no args from inside a git working tree whose origin is on GitHub. Findings are harvested from vulnhunter-labeled GitHub issues on that repo. The full report is staged from the publish repo named by the vulnhunt-results-dir marker. Fixes happen on per-finding worktrees rooted at <repo>/.vulnhunter-fix/. Delivery pushes branches and opens PRs back to the same repo; the source issue is commented and closed. |
See prompts/parse_issues.md, then plan.md → implement.md → verify.md → deliver.md (in-place section). |
| Fork (legacy / cross-org) | User passes TARGET_REPO + RESULTS_PATH explicitly. Skill clones into ./work/, forks the target into a configured org, and delivers PRs to the fork. |
See prompts/parse.md, then the same downstream phases (fork section in deliver.md). |
Mode dispatch (mandatory)
Before any other action, decide the mode. Run the canonical dispatcher script and parse its stdout:
# Returns one of: mode=in_place / mode=fork / mode=ambiguous / mode=none
# Exit 0 for in_place/fork/none; exit 2 for ambiguous (caller must resolve).
DISPATCH="$(bash "${SKILL_DIR}/scripts/detect_mode.sh" \
"${TARGET_REPO:-}" "${RESULTS_PATH:-}")"
echo "$DISPATCH"
The script is the source of truth for the dispatch rule — it has its own tests under tests/test_detect_mode_sh.py. Behavior summary:
What ships with it
60 files 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.
- collaborators.json 93 B
- config.json 2.1 KB
- evals/evals.json 3.3 KB
- prompts/agent_fix.md 1.7 KB
- prompts/deliver.md 31 KB
- prompts/implement.md 38 KB
- prompts/parse_issues.md 37 KB
- prompts/parse.md 2.6 KB
- prompts/plan_fork.md 3.3 KB
- prompts/plan.md 8.3 KB
- prompts/reviewer_test.md 2.6 KB
- prompts/sweep.md 2.6 KB
- prompts/tier_judgment.md 4.2 KB
- prompts/verify.md 27 KB
- prompts/worker_agent_authz.md 2.2 KB
- prompts/worker_agent_common.md 25 KB
- prompts/worker_agent_config.md 2.8 KB
- prompts/worker_agent_crypto.md 2.7 KB
- prompts/worker_agent_injection.md 2.3 KB
- prompts/worker_agent_resource.md 2.0 KB
- pyproject.toml 2.4 KB
- README.md 4.9 KB
- references/anti-merge-math.md 2.2 KB
- references/approved-crypto-algorithms.yaml 2.4 KB
- references/approved-key-sources.yaml 2.1 KB
- references/committed-test-naming-rule.md 2.0 KB
- references/cwe-fix-patterns.md 6.8 KB
- references/finding-schema.json 2.2 KB
- references/fix_plan-schema.json 4.5 KB
- references/fix-completeness-rubric.md 4.5 KB
- references/idempotency-key-rule.md 1.4 KB
- references/README.md 3.4 KB
- references/remediation-rigor.md 9.2 KB
- references/repo-type-adapters.md 3.8 KB
- references/residual-risk-rules.md 3.3 KB
- references/result-schema.json 10 KB
- references/severity-mask-rule.md 2.0 KB
- references/sweep-algorithm.md 3.3 KB
- references/sweep-patterns.md 2.9 KB
- references/test-quality-rubric.md 3.6 KB
- references/triage-schema.json 3.9 KB
- references/verification-table-rules.md 3.7 KB
- scripts/_skill_bootstrap.py 4.1 KB runs code
- scripts/anti-merge-check.py 3.2 KB runs code
- scripts/build_graph.py 6.9 KB runs code
- scripts/check_repo_access.sh 943 B runs code
- scripts/check-body-completeness.py 5.5 KB runs code
- scripts/check-committed-test-naming.py 7.0 KB runs code
- scripts/check-idempotency.py 1.5 KB runs code
- scripts/check-scope.py 3.5 KB runs code
- scripts/check-severity-mask.py 4.2 KB runs code
- scripts/clone_repo.sh 3.8 KB runs code
- scripts/cluster_score.py 5.8 KB runs code
- scripts/compute-completeness-tier.py 9.6 KB runs code
- scripts/crypto-trust-chain-checkers.py 15 KB runs code
- scripts/detect_mode.sh 4.3 KB runs code
- scripts/heading-sync-lint.py 7.4 KB runs code
- scripts/issue_intake.py 7.1 KB runs code
- scripts/language-detect.py 4.5 KB runs code
- scripts/parse_results.py 13 KB runs code
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.
- 2mo ago First seen · 407 lines · 135 tokens per session scan A 836f15939203
vulnhunter-fix is a skill published in the GitHub repository capitalone/VulnHunter (1,134 stars, last pushed 21d ago), licensed Apache-2.0. It adds 135 tokens to every session and 7,866 once invoked, about $0.0005 per session on Opus 5.5. 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 skills, from other repositories
engram-testing-coverage
TDD and coverage standards for Engram. Trigger: When implementing behavior changes in any package.
principle-test-behavior-not-implementation
Apply when you write, change, or keep a test. Assert what the code's users can observe, so that the test fails when that behavior breaks and keeps passing when only the internals change.
conductor-implement
Execute tasks from a track's implementation plan following TDD workflow.
tdd
This skill should be used when the user wants to implement features or fix bugs using test-driven development. Enforces the RED-GREEN-REFACTOR cycle with vertical slicing, context isolation between test writing and implementation, human checkpoints, and auto-test feedback loops. Uses multi-agent orchestration with the…
mobiai-mobile-tdd
You MUST use this before writing any implementation code for a mobile feature, bug fix, refactor, or behavior change. Tests come before implementation — no exceptions.
iterative-development
TDD iteration loops using Claude Code Stop hooks - runs tests after each response, feeds failures back automatically.