Borrowing it
Nothing to install: this file belongs to MannanSaood/termi. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/MannanSaood/termi/main/.cursor/skills/audit-unwraps/SKILL.mdgit clone --depth 1 https://github.com/MannanSaood/termiWrote 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/mannansaood/termi/audit-unwraps)<a href="https://agentmods.dev/skills/mannansaood/termi/audit-unwraps"><img src="https://agentmods.dev/badge/skills/mannansaood/termi/audit-unwraps/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/mannansaood/termi/audit-unwraps"><img src="https://agentmods.dev/badge/skills/mannansaood/termi/audit-unwraps.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 | $0.00048 | $0.00795 |
| Opus 5 | $0.00024 | $0.00398 |
| Sonnet 5 | $0.00010 | $0.00159 |
| Haiku 4.5 | $0.00005 | $0.00080 |
Grade A, and why
audit-unwraps 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 8d 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 — 72 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Use this if clippy::unwrap_used flags something and it's not obvious how
to fix it, or when doing a periodic safety sweep.
Step 1: find them, excluding test code
cd rust/src
grep -rn "\.unwrap()\|\.expect(" --include="*.rs" . | grep -v "^\./.*test"
This is a rough filter — always manually check whether a hit is actually
inside a #[cfg(test)] mod tests { ... } block before treating it as a
production issue. .unwrap() in test assertions is normal and expected;
the project's mod tests blocks are explicitly annotated with
#[allow(clippy::unwrap_used, clippy::expect_used)] for exactly this
reason — don't "fix" those.
Step 2: classify each production hit
For each real (non-test) occurrence, ask:
-
Is this a
Mutex::lock()? → useutils::sync_ext::LockExt::lock_safe()instead. See that module's doc comment for the one case where this is NOT appropriate (mutexes protecting real invariants where a torn write could leave data in a genuinely broken state — rare in this codebase, but check). -
Is this parsing/converting something that could legitimately fail (user input, file content, JNI-provided data)? → propagate via
?using the appropriate error type fromutils::error(VfsError,PtyError,TerminalError,JniErrorCode). If the calling function doesn't return aResultyet, that's the actual fix — change its signature, don't swallow the error with a fallback default silently (silent fallbacks hide real problems just as much as a panic does, they just do it later and more confusingly). -
Is this "genuinely can't happen" (e.g. unwrapping a regex you just compiled from a string literal)? → still avoid
.unwrap(). Either:- Use
.expect("clear message about why this can't fail")— this is marginally better since it documents intent, butclippy::expect_usedis enforced too, so this still needs a#[allow(clippy::expect_used)]with a comment justifying it, used sparingly and only for genuine invariants, not convenience. - Better: restructure so the "can't happen" case is checked once at startup/construction time and stored as an already-valid value, rather than re-asserted on every call.
- Use
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.
- 8d ago First seen · 72 lines · 48 tokens per session scan A b84caa385a76
audit-unwraps is a skill published in the GitHub repository MannanSaood/termi (8 stars, last pushed 9d ago), licensed MIT. It adds 48 tokens to every session and 795 once invoked, about $0.0002 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 2026-08-31.
Other skills, from other repositories
shepherd
Use when asked to shepherd, babysit, monitor, or poll open pull requests or merge requests, including triaging review feedback, CI failures, and routine follow-up.
kotlin-api-design
Use when designing or reviewing Kotlin function ownership, member or extension functions, factories, single-field domain types, value classes, data classes, Kotlin Multiplatform expect/actual declarations, or platform service boundaries.
kotlin-concurrency-and-flow
Use when writing or reviewing Kotlin coroutine scope ownership, raw Thread or Executor work, init launches, non-suspending launch APIs, runBlocking, cancellation, StateFlow, SharedFlow, Channel, stateIn, SharingStarted, state updates, or one-shot events.
shepherd
Use when asked to shepherd, babysit, monitor, or poll open pull requests or merge requests — including triaging review comments, detecting CI failures, fixing trivial CI issues, and keeping PRs/MRs moving without manual intervention.
derive-the-flag-dont-store-and-correct-it
A boolean that is a pure function of state already being collected gets stored as its own mutableStateOf anyway, seeded with a guess and corrected a frame later by a LaunchedEffect — so the first frame renders the guess, and later changing only the seed value does nothing once rememberSaveable has already saved the…
a-stated-rule-needs-annotated-exceptions
A design rule with legitimate exceptions survives only if every exception carries its reason at the call site and the rule itself is greppable — otherwise nothing distinguishes an exception from a violation and the rule silently rots. Covers where the rule statement goes, where the reasons go, scoping the audit to the…