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 Sassy-Dog/sassydog-skills --skill sentry-triagegit clone --depth 1 https://github.com/Sassy-Dog/sassydog-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/sassy-dog/sassydog-skills/sentry-triage)<a href="https://agentmods.dev/skills/sassy-dog/sassydog-skills/sentry-triage"><img src="https://agentmods.dev/badge/skills/sassy-dog/sassydog-skills/sentry-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/sassy-dog/sassydog-skills/sentry-triage"><img src="https://agentmods.dev/badge/skills/sassy-dog/sassydog-skills/sentry-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 | $0.00153 | $0.01669 |
| Opus 5 | $0.00077 | $0.00834 |
| Sonnet 5 | $0.00031 | $0.00334 |
| Haiku 4.5 | $0.00015 | $0.00167 |
Grade A, and why
sentry-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 7d 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 — 65 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Sentry Triage
Gate-and-escalate triage: pull not-resolved Sentry issues for a project, apply a qualifying gate, cross-reference against GitHub, and (only with explicit approval) escalate survivors to GitHub issues. Reads only — this skill never mutates Sentry state.
Inputs to establish first
- Org + project slug(s) — from the caller or by listing projects. Don't hunt for phantom projects: platforms sharing a DSN don't appear separately.
- Gate policy — the caller's thresholds, or the defaults in
references/qualifying-gate.md. - Escalation target — repo (and optional board IDs) if filing is on the table; otherwise this is a read-only report.
Workflow
1. Pull (discovery only)
MCP-first: use the connected Sentry MCP server's project-listing and issue-search tools (resolve the project slug first, then search). Do not hardcode mcp__... tool IDs — resolve by capability. No MCP connected → references/api-fallback.md (REST with SENTRY_AUTH_TOKEN).
Query !is:resolved — one call covers unresolved + ignored (status is exclusive; boolean OR returns HTTP 400 — full syntax notes in references/query-syntax.md). Pull per issue: shortId, title, count, userCount, lastSeen, firstSeen, level, culprit, permalink, status.
Issue search is a DISCOVERY tool, never a SCORING tool. It tells you which fingerprints exist; its counts must never reach a gate or a plate. The MCP's search_issues renders Events and Users scoped to its period argument, labelled identically to lifetime totals and with nothing in the output marking them as windowed — so a long-running issue reads as a one-off. Measured on Sassy-Dog/velovate, 2026-08-20 and re-verified live 2026-08-21: VELOVATE-WORKERS-2 returned 1 event / 1 user against a true 30 occurrences / 3 users, a 30× undercount, and the 30d pull returned 1 event too — widening period is not a workaround (issue #218). firstSeen is rescaled the same way: search reported it as 2 days ago against a true first-seen of 2026-06-25, 57 days earlier, so a long-running fingerprint reads as a fresh regression from the latest deploy. On the true numbers that issue qualifies (production environment, 3 users ≥ 2); on the windowed pair it fails every clause of the severity rule. The gate outcome inverts, silently. The same names mean different things per transport, and REST is the one that is already lifetime; the table in references/api-fallback.md is the authority for both.
What ships with it
3 files 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.
- 7d ago Changed 3dd0fe3161f2
- 12d ago First seen · 65 lines · 153 tokens per session scan A fb2db37d30fe
sentry-triage is a skill published in the GitHub repository Sassy-Dog/sassydog-skills (0 stars, last pushed 3d ago), licensed Apache-2.0. It adds 153 tokens to every session and 1,669 once invoked, about $0.0008 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
browse-flows
Browse Power Automate environments and flows interactively. Use when the user wants to browse, list, or explore their flows and environments.
operating-cadence
Designs the rhythm an organization runs on — which reviews happen weekly, monthly and quarterly, what each one decides, who owns the numbers presented, and how a signal at the front line reaches the people who can act on it. Use this to set up a management operating system, fix a meeting calendar that produces no…
propose-issue
Use when the user asks to propose an issue, file or open a bug report, or turn a reported defect into a tracked issue. Not for fixing the defect, triaging existing issues, or reviewing PRs.
defect
A command for registering a software defect, meaning behavior that does not work as intended, and creating a task to fix it. The defect and its fix task receive tracked IDs and are stored in the project's backlog and task folders.
workflow-audit-emit-issues
After a codebase audit, groups audit findings by subsystem (path prefix) and emits them as context-rich GitHub issues with template discovery, label selection, and a batch preview-confirm gate before filing. Keywords — audit findings grouped github issues subsystem emit.
upstream-tracker
Manage upstream issue tracking for this project. Use when the user wants to log a bug or friction point in a vendor package or npm dependency, review open upstream items, resolve a tracked issue, run a trend review, generate the upstream observations section of a sprint retrospective, promote upstream observations to…