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 robzilla1738/roberts-skills --skill triagegit clone --depth 1 https://github.com/robzilla1738/roberts-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/robzilla1738/roberts-skills/triage)<a href="https://agentmods.dev/skills/robzilla1738/roberts-skills/triage"><img src="https://agentmods.dev/badge/skills/robzilla1738/roberts-skills/triage/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/robzilla1738/roberts-skills/triage"><img src="https://agentmods.dev/badge/skills/robzilla1738/roberts-skills/triage.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 | $0.00082 | $0.02973 |
| Opus 5 | $0.00041 | $0.01486 |
| Sonnet 5 | $0.00016 | $0.00595 |
| Haiku 4.5 | $0.00008 | $0.00297 |
Grade A, and why
triage 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 9d 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 — 255 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Triage & Repro
Companion skill that converts raw suspicion into a precise, ranked, evidence-backed report. It owns the finding schema, the severity × confidence model, and the discipline of building a minimal reproduction before anything is called Confirmed.
Use it standalone on a single suspected bug, or as the scoring/reporting stage of bughunt.
First principle: evidence or it didn't happen
A finding is a claim that specific code does the wrong thing under a specific condition. Every finding must carry:
- Location —
file:line(the exact site, not "somewhere in this module"). - Trace — the data/control path from cause to effect.
- Trigger — the concrete condition or input that makes it fire.
- Impact — what actually goes wrong (crash, corruption, wrong result, leak, exposure).
If you cannot state all four, the item is a question, not a finding. Demote it or investigate further. A short list of real bugs beats a long list of maybes — noise destroys trust in the report.
Severity rubric
Rate impact if triggered, independent of how likely it is.
| Severity | Meaning |
|---|---|
| Critical | Data loss/corruption, RCE, auth bypass, secret disclosure, crash on common path |
| High | Wrong result users rely on, privilege issue, leak/exhaustion under normal load, crash on a real edge |
| Medium | Incorrect behavior on an uncommon-but-reachable path; degraded reliability; recoverable |
| Low | Minor correctness/robustness issue, narrow edge, cosmetic-but-wrong |
Impact rubric (what "severity" means per finding class)
bughunt hunts more than security bugs. A finding's impactClass tells you which yardstick
to hold its severity against, so a pain-point or inefficiency isn't force-fit into a
security frame (and isn't dismissed because it isn't a CVE). Same four severity levels,
calibrated per class:
| impactClass | What Critical/High looks like here | What Low looks like |
|---|---|---|
| security | auth bypass, RCE, secret/PII disclosure, injection on a reachable path | theoretical issue behind strong guards |
| correctness | wrong result users act on; silent data divergence | wrong only on a contrived edge |
| reliability | crash/hang/outage under normal load; fail-open | degrades only under rare conditions |
| performance | O(n²)/N+1/unbounded growth that breaks at real scale | constant-factor waste on a cold path |
| data-integrity | destructive/irreversible migration, corruption, lost writes | recoverable, narrow-window skew |
| dx | broken build/test, footgun that routinely costs dev hours | stale TODO, cosmetic friction |
| ux | user can't recover (white screen, silent failure, double-charge) | minor polish / missing affordance |
| supply-chain | known-vulnerable/typosquat dep on a reachable path | unpinned but low-risk dev dep |
What ships with it
1 file 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.
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.
- 9d ago First seen · 255 lines · 82 tokens per session scan A f59255be479d
triage is a skill published in the GitHub repository robzilla1738/roberts-skills (5 stars, last pushed 3mo ago), licensed MIT. It adds 82 tokens to every session and 2,973 once invoked, about $0.0004 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
material-ui-nextjs
Integrates Material UI with Next.js App and Pages routers using @mui/material-nextjs, Emotion cache providers, next/font, CSS layers with Tailwind/CSS Modules, Link component prop patterns, CSS theme variables SSR notes, and App Router useSearchParams + Suspense. Use when setting up or debugging MUI in a Next.js app.
find-bugs
Investigate reported or suspected bugs in this daisyUI monorepo, including proactive bug searches in packages/daisyui and packages/docs. Use when Codex must reproduce a defect, prove product impact, identify its root cause and isolation boundary, evaluate non-code solution options, or prepare a decision-ready bug-fix…
chromatic-diff
A diagnostic workflow for explaining why a Chromatic visual test changed. Chromatic is a service that compares screenshots of UI stories across builds.
test-fix
Diagnose and minimally fix a failing test, then re-run.
ios-ettrace-profiler
Before invoking Apple-only binaries, confirm the execution context is macOS. From Windows or Linux, run those steps in a Mac SSH project or through an already configured remote transport; do not retry missing Apple binaries locally.
macos-runtime-debugger
Before invoking Apple-only binaries, confirm the execution context is macOS. From Windows or Linux, run those steps in a Mac SSH project or through an already configured remote transport; do not retry missing Apple binaries locally.