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 extracting-nday-from-a-patchgit 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/extracting-nday-from-a-patch)<a href="https://agentmods.dev/skills/unboundcompute/security-agent-skills/extracting-nday-from-a-patch"><img src="https://agentmods.dev/badge/skills/unboundcompute/security-agent-skills/extracting-nday-from-a-patch/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/extracting-nday-from-a-patch"><img src="https://agentmods.dev/badge/skills/unboundcompute/security-agent-skills/extracting-nday-from-a-patch.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 | $0.00121 | $0.01527 |
| Opus 5.5 | $0.00048 | $0.00611 |
| Sonnet 5.5 | $0.00024 | $0.00305 |
| Haiku 4.5 | $0.00012 | $0.00153 |
Grade A, and why
extracting-nday-from-a-patch 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.
Extracting n-day from a patch: the fix is the map
A security fix is a confession. It tells you exactly where the bug was, what the missing check should have been, and, by omission, which other paths the author did not think to cover. Reading a patch backward, from fix to bug, is one of the highest-yield techniques in vulnerability research: the hard part (locating the defect) is already done for you, and the incomplete fixes are waiting.
When to use
- A fix commit or security release landed and you want to know what it addressed.
- An advisory is deliberately vague but links a diff, a PR, or a tag.
- You maintain or depend on a fork and need to know if a fix upstream applies.
- You confirmed one bug via its patch and want the instances the patch missed.
Scope check
You may read and reason about public patches and advisories freely. You may only test or exploit against code and systems you are authorized for (your own, OSS you run, CTF, in-scope engagement). Analysis is not authorization to attack.
The loop
-
Read the fix, infer the bug. Look at what the patch adds or tightens, and name the vulnerability it implies. A new length or bounds check implies an out-of-bounds read or write. A new authorization call implies missing access control. A new escape or parameterization implies injection. A new normalization implies traversal or confusion. The added guard tells you the invariant that was being violated.
-
Locate the pre-patch sink and its precondition. In the vulnerable revision, find the exact operation that was unsafe and the condition under which it misbehaves (the input shape, size, or state the new guard now rejects). This is the bug, stated concretely.
-
Reconstruct the reachable path. Confirm an untrusted input could reach that sink in the vulnerable version. Because the fix exists, the bug is real by construction; your job is to establish the trigger, not to re-prove the flaw.
-
Test the fix for completeness. This is the payoff. Ask whether the new guard sits on every path to the sink or only the reported one. Common gaps: the check was added to one caller but a second caller still reaches the sink raw; the check runs after an early operation that is already unsafe; the guard covers one input shape but a sibling shape (a different encoding, a second parameter) slips past. An incomplete fix is a live bug in a patched version.
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 · 121 tokens per session scan A ffc00edbdf4c
extracting-nday-from-a-patch is a skill published in the GitHub repository UnboundCompute/security-agent-skills (5 stars, last pushed 1mo ago), licensed MIT. It adds 121 tokens to every session and 1,527 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
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.