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 detecting-memory-safety-bugsgit 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/detecting-memory-safety-bugs)<a href="https://agentmods.dev/skills/unboundcompute/security-agent-skills/detecting-memory-safety-bugs"><img src="https://agentmods.dev/badge/skills/unboundcompute/security-agent-skills/detecting-memory-safety-bugs/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/detecting-memory-safety-bugs"><img src="https://agentmods.dev/badge/skills/unboundcompute/security-agent-skills/detecting-memory-safety-bugs.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 | $0.00124 | $0.01428 |
| Opus 5 | $0.00062 | $0.00714 |
| Sonnet 5 | $0.00025 | $0.00286 |
| Haiku 4.5 | $0.00012 | $0.00143 |
Grade A, and why
detecting-memory-safety-bugs 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 4d 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 — 116 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Detecting memory-safety bugs
Temporal and spatial memory bugs are the ones a sink-based catalog usually doesn't model: there's no single "dangerous function" to grep - the bug is a relationship between a pointer's lifetime and its use, or between an index and a bound, spread across code paths. So you hunt them by reasoning about lifetime and bounds, not by matching a call. This skill covers the five workhorse classes and how to confirm each.
When to use
- The target is C/C++/Rust-unsafe/CGo or any unmanaged memory, and you want the
classes a catalog leaves out (
hunting-bugs-with-a-code-graphflags these as out-of-catalog and sends you here). - You're reviewing allocators, parsers, serializers, ring buffers, refcounting, or anything doing pointer arithmetic.
Scope check
Authorized source only. If you can't name the authorization, stop.
The five classes and how to confirm each
For every candidate, the confirmation is a path: an allocation/definition site, the operation that changes its state, and the use - read the source at each.
-
Use-after-free (UAF). A pointer is used after its object is freed. Hunt: for each
free/delete/refcount-drop, ask what still holds this pointer and can any path reach a use after this point - including aliases stored in structs, callbacks, and error paths. Confirm: a live path free → … → deref with no reassignment in between. Watch the classic shapes: free-in-a-loop then use, free in an error branch then fallthrough use, and a cached pointer that outlives the object. -
Double-free. The same allocation freed twice. Hunt: two frees reachable in sequence on some path without a nulling/reassignment between them; ownership handed to a callee that also frees; error paths that free then fall into a cleanup that frees again. Confirm: both frees hit the same object on one path.
-
Out-of-bounds read/write. An index or pointer crosses a buffer's bound. Hunt: every buffer access where the index/length is attacker-influenced or derived from a separate field (length prefixes are a hotspot). Confirm: trace the index/length source and the buffer's true size; the bug is any path where
index/lencan exceedsize- off-by-one at<= size, uncheckedmemcpylength, integer-truncated length, signed/unsigned confusion at the check.
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.
- 4d ago First seen · 116 lines · 124 tokens per session scan A 20a3447e01f8
detecting-memory-safety-bugs is a skill published in the GitHub repository UnboundCompute/security-agent-skills (5 stars, last pushed yesterday), licensed MIT. It adds 124 tokens to every session and 1,428 once invoked, about $0.0006 per session on Opus 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
ffuf-claude-skill
Skill "ffuf-claude-skill" from Agent-Threat-Rule/agent-threat-rules, covering ffuf claude skill, when to use this skill and instructions.
codeinspectus-fix-one
Investigate and remediate exactly one user-selected CodeInspectus finding with evidence-gated reproduction, a separately approved minimal patch, focused regression testing, and an exact-prior-scan rescan. Use when a user asks an agent to examine, reproduce, fix, or verify one CodeInspectus finding without batching…
gauntlex:doctor
Full GAUNTLEX environment health check — model reachable, ChromaDB writable, AVF fixtures pass, no unexpected outbound network (air-gap verification). Use before first run or when debugging failures.
prodcheck-review
Review this codebase against the prodcheck pre-production checklists — security, performance, scale, integrations and post-launch readiness. 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 citations…
diagnose-bundle
Maintainer/dev skill that triages an anonymised diagnostic bundle (appsec-diag-.tgz, produced by scripts/diagnosticbundle.py) a user sent after a pipeline failure. Runs the deterministic inspect, then cross-references the plugin source (scripts/, compact runtimes, agents/, AGENTS.md) and known-bug history to produce a…
diagnose-run
Find the plugin defects behind the issues a create-threat-model run recorded. Reads .run-issues.json, re-reads each symptom next to the plugin's own code, and reports a file:line root cause per issue, or classifies it as an environment or expected condition. Read-only against the plugin; writes only .run-bugs.json.…