Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/UnboundCompute/security-agent-skillsnpx agentmods add skills/unboundcompute/security-agent-skills/writing-vuln-reportsWrote 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/writing-vuln-reports)<a href="https://agentmods.dev/skills/unboundcompute/security-agent-skills/writing-vuln-reports"><img src="https://agentmods.dev/badge/skills/unboundcompute/security-agent-skills/writing-vuln-reports/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/writing-vuln-reports"><img src="https://agentmods.dev/badge/skills/unboundcompute/security-agent-skills/writing-vuln-reports.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 | $0.00105 | $0.01508 |
| Opus 5.5 | $0.00042 | $0.00603 |
| Sonnet 5.5 | $0.00021 | $0.00302 |
| Haiku 4.5 | $0.00011 | $0.00151 |
Grade A, and why
writing-vuln-reports 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 — 123 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Writing vulnerability reports
A finding that isn't clearly reported doesn't get fixed. The job of a report is to move a specific defect from your head into the maintainer's, with enough proof that they can reproduce it in minutes and enough impact that they prioritize it, and nothing else. This skill turns a schema-shaped finding into that report.
When to use
- A finding is
confirmed(see FINDING-SCHEMA.md) and you're writing it up for a human: bounty submission, advisory, ticket, email. - You have several findings and need consistent, triager-friendly writeups.
Scope check
Only report findings from authorized testing, to the party entitled to receive them (the program's channel, the maintainer's security contact, your client). Don't disclose someone else's data or a third party's system you weren't scoped to touch. When in doubt about the channel, ask before sending - a report is outward-facing and hard to unsend.
The structure that gets findings fixed
Lead with impact, prove it fast, make the fix obvious. Sections:
- Title - the specific defect and where. "IDOR in
admin_bulk_deletelets any user delete any post," not "Access control issue." - Summary (2 to 3 sentences) - what the bug is, who can trigger it, what they get. A busy triager should grasp severity from this alone.
- Impact - the concrete consequence, tied to the target's threat model. Who is harmed, what they lose, what precondition is needed (auth level, timing, config). Don't inflate; don't undersell.
- Steps to reproduce - numbered, exact, copy-pasteable. Real request(s) with method, path, headers that matter, and body. State the starting state (logged in as user A, two accounts, a seeded record). A stranger must reproduce it from these steps alone.
- Proof - the request/response pair (or crash/log) that shows it firing.
Redact secrets and unrelated PII. For whitebox findings, include the
source→sink path and the decisive code excerpt from the schema's
evidence. - Root cause - the actual defect, at the decisive hop (the missing check, the unbounded copy, the non-atomic act). One or two sentences.
- Remediation - the fix at that hop, plus the general pattern ("parameterize the query," "resolve against base and reject escapes," "atomic conditional update"). Offer the fix; don't demand a specific patch.
- Severity - a rating with the reasoning. If the program uses CVSS, give the vector and score; otherwise a justified low/med/high/critical. The reasoning matters more than the number.
- References (optional) - CWE id, relevant advisory, similar prior fixes.
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 · 123 lines · 105 tokens per session scan A 7f35a8de0592
writing-vuln-reports is a skill published in the GitHub repository UnboundCompute/security-agent-skills (5 stars, last pushed 1mo ago), licensed MIT. It adds 105 tokens to every session and 1,508 once invoked, about $0.0004 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
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…
integrate-arcjet-guard-genkit
Integrate Arcjet security into a Genkit JS agent using @arcjet/guard — wrap ai.defineTool, put guardMiddleware on generate({ use }) for unwrapped / MCP / filesystem tools, and read a caller-owned id from generate({ context }). Use when asked to add Arcjet to genkit, rate limit its tools, screen inbound messages, or…
integrate-arcjet-guard-tanstack-ai
Integrate Arcjet security into a TanStack AI chat() app using @arcjet/guard — put guardMiddleware first on chat({ middleware }) so onBeforeToolCall gates tools, and read a caller-owned id from chat({ context }). Use when asked to add Arcjet to TanStack AI, rate limit its tools, screen inbound messages, or block prompt…
integrate-arcjet-guard-claude-managed-agents
Integrate Arcjet security into Claude Managed Agents (hosted REST+SSE, beta managed-agents-2026-04-01) using @arcjet/guard — screen user.message / initialevents before sessions.events.send, and gate custom tools on agent.customtooluse. Use when asked to add Arcjet to Claude Managed Agents, rate limit custom tools, or…
linear-claude-skill
Manage Linear issues, projects, and teams.
varlock-claude-skill
Secure environment variable management ensuring secrets are never exposed in Claude sessions, terminals, logs, or git commits.