Skip to content
Back to skills

Sentry

ASecurity

Check Sentry for unresolved issues and run triage on demand. Triggered by requests to check Sentry errors, run Sentry triage, or handle Sentry issues.

  • 26 stars
  • 0 votes
  • 0 copies
  • 5 views
  • Added September 2, 2026
ai-agentspythonbashgit

Security analysis

A100/100

Scanned September 2, 2026

npx -y skills add tomcounsell/ai --skill sentry --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Sentry?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Sentry
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/tomcounsell-sentry/badge)](https://www.skillsdirectory.com/skills/tomcounsell-sentry)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: sentry
description: "Check Sentry for unresolved issues and run triage on demand. Triggered by requests to check Sentry errors, run Sentry triage, or handle Sentry issues."
allowed-tools: Bash
user-invocable: true
argument-hint: "[--apply]"
---

# Sentry Triage

Run the Sentry triage pipeline on demand. Classifies all unresolved issues (A–E), files GitHub issues for actionable bugs, and sends a Telegram summary.

## Classification

| Class | Label      | Action                                     |
|-------|------------|--------------------------------------------|
| A     | Noise      | Test/mock/harness errors → Sentry ignored  |
| B     | Transient  | Rate limits, network errors → archived     |
| C     | Actionable | Real bugs (≥10 events) → GitHub issue      |
| D     | Review     | Ambiguous → listed for human review        |
| E     | Stale      | No events in 30 days → Sentry resolved     |

## Apply Mode

By default the triage runs **dry-run** — it classifies and reports but does not file GitHub issues or change Sentry state.

Pass `--apply` to enable live writes:

```
/sentry --apply
```

## How to Run

The `SENTRY_TRIAGE_APPLY` env var controls live writes (default `0` = dry-run;
`1` = files GitHub issues + updates Sentry state):

```bash
cd ~/src/ai && python -c "
from reflections.sentry_triage import run_sentry_triage
result = run_sentry_triage()
print(result['summary'])
if result['status'] != 'disabled':
    show = False
    for line in result.get('findings', []):
        if line.startswith('Class C') or line.startswith('Class D'):
            show = True
        elif line.startswith('Class '):
            show = False
        if show:
            print(line)
"
```

For live mode, run the same command prefixed with `SENTRY_TRIAGE_APPLY=1`.

## Argument Handling

Check the user's invocation for `--apply`. If present, use the live mode command. Otherwise use dry-run.

If the user says "apply changes", "file the issues", "do it for real", or similar — first run the dry-run to show what will be actioned, state the count explicitly ("This will file N GitHub issues and update M Sentry issues"), then ask for confirmation before running live mode.

## After Running

- Show the summary line (counts by class).
- Show all Class C items — these are the actionable bugs.
- Show Class D items — note that only the top 5 are listed even if more exist.
- Omit Class A/B/E details unless the user asks.

If running dry-run and Class C issues exist, offer: "Run `/sentry --apply` to file GitHub issues for the N actionable bugs."

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.