Auth Failure Log Triage
Investigate reported auth, credential, permission, API proxy, 401, 403, 503, OAuth, token, or secret-resolution failures by querying VictoriaLogs first. Use when a user says an integration auth failed, a tool got unauthorized/forbidden/service unavailable, an api-proxy request failed, or asks why credentials/secrets are not working.
Get the complete skill folder, including its scripts and reference files.
When to use this skill
Use Auth Failure Log Triage when an AI agent needs a reusable SKILL.md workflow for this job: Investigate reported auth, credential, permission, API proxy, 401, 403, 503, OAuth, token, or secret-resolution failures by querying VictoriaLogs first. Use when a user says an integration auth failed, a tool got unauthorized/forbidden/service unavailable, an api-proxy request failed, or asks why credentials/secrets are not working.
When not to use it
Skip Auth Failure Log Triage when the task is outside the coding category, or when a more specific skill in this directory already covers the same workflow with clearer triggers.
How to install
Where will you use it?
mkdir -p ~/.claude/skills/auth-failure-log-triage # Save SKILL.md to ~/.claude/skills/auth-failure-log-triage/SKILL.md
- Claude Code personal install: place the skill and its bundled files at ~/.claude/skills/auth-failure-log-triage/SKILL.md. For a shared project, commit the folder at .claude/skills/auth-failure-log-triage/ instead.
- Current Claude Code detects SKILL.md changes in watched directories during the session. If you just created a top-level skills directory, run /reload-skills. Check that the skill is listed before testing its trigger.
- For Claude chat or Cowork in the Desktop app, manage uploaded skills in Customize > Skills; copying files into a local Claude Code folder is not a Desktop install. Other agents may use different install locations.
Full install guide for Claude, Cursor, and Codex →
Try a starter prompt
After installing the skill, paste this starter prompt into your agent to verify it triggers on a small task.
Use the Auth Failure Log Triage skill for this task: Confirm the affected surface from the user message: tool/integration name, URL host, thread key, execution id, pod/service, and rough time window. Confirm any required inputs or files before making changes.
What supports this recommendation?
83/100 provisional content score. This measures documentation, practical detail and source signals. It is not a measured task success rate or a security certification.
- Documentation
- 92/100
- Practical detail
- 92/100
- Source evidence
- 63/100
- Source signals
- 72/100
Model compatibility evidence
- claudeuntested
No model-specific signal or recorded compatibility test was found.
- chatgptuntested
No model-specific signal or recorded compatibility test was found.
- geminiuntested
No model-specific signal or recorded compatibility test was found.
- copilotuntested
No model-specific signal or recorded compatibility test was found.
- llamauntested
No model-specific signal or recorded compatibility test was found.
- perplexityuntested
No model-specific signal or recorded compatibility test was found.
- mistraluntested
No model-specific signal or recorded compatibility test was found.
- grokuntested
No model-specific signal or recorded compatibility test was found.
Inferred compatibility is based on content signals. Check the app, its available tools and setup requirements before running a skill.
What this skill does
# Auth Failure Log Triage
When a user reports an auth, permission, credential, API proxy, OAuth, token, secret, `401`, `403`, or `503` failure, inspect live runtime evidence before proposing secret or permission changes.
## Workflow
1. Confirm the affected surface from the user message: tool/integration name, URL host, thread key, execution id, pod/service, and rough time window. 2. Discover `vlogs` if needed:
```bash call discover vlogs ```
3. Query likely auth failures in the last 24h, narrowing when the user gave a host, pod, thread, or execution id.
For Centaur API proxy / Iron Proxy issues:
```bash call vlogs query '{"query":"kubernetes.pod_name:centaur-api-proxy-* AND (audit.status_code:401 OR audit.status_code:403 OR audit.status_code:503 OR \"failed to fetch secret\" OR unauthorized OR forbidden OR auth OR oauth OR token OR secret)","start":"24h","limit":100}' ```
For a specific thread:
```bash call vlogs thread_trace '{"thread_key":"THREAD_KEY","start":"24h","limit":500}' call vlogs tool_usage_by_thread '{"thread_key":"THREAD_KEY","start":"24h","limit":200}' ```
For a specific execution:
```bash call vlogs execution_timeline '{"execution_id":"EXECUTION_ID"}' ```
4. If the direct query is empty, find labels first:
```bash call vlogs field_values '{"field":"service","query":"*proxy*","start":"24h","limit":200}' call vlogs field_names '{"query":"service:iron-proxy","start":"24h"}' call vlogs streams '{"query":"proxy","start":"24h"}' ```
5. Summarize only what the logs support:
- Counts by status code: `401`, `403`, `503`. - Secret-resolution errors and the secret reference name, without exposing secret values. - Timestamp, pod, service, host, method, path, and status for concrete failures. - Whether auth transforms injected credentials, if shown by proxy fields such as `request_transforms`. - Distinguish upstream denial (`audit.status_code=403`) from local proxy/secret resolution failure (`failed to fetch secret`).
## Output Rules
- Lead with whether log evidence was found. - Do not claim a secret or permission is wrong unless logs show that. - Do not recommend secret rewiring until you have checked live `vlogs` evidence and compared the relevant error shape. - If no logs are found, say which query/window was checked and ask for the missing discriminator: time, thread key, execution id, host, or integration name.
Intended uses
- Confirm the affected surface from the user message: tool/integration name, URL host, thread key, execution id, pod/service, and rough time window.
- Discover vlogs if needed:
- Query likely auth failures in the last 24h, narrowing when the user gave a host, pod, thread, or execution id.
- If the direct query is empty, find labels first:
- Summarize only what the logs support:
- Counts by status code: 401, 403, 503.
- Secret-resolution errors and the secret reference name, without exposing secret values.
- Timestamp, pod, service, host, method, path, and status for concrete failures.
- Whether auth transforms injected credentials, if shown by proxy fields such as requesttransforms.
- Distinguish upstream denial (audit.statuscode=403) from local proxy/secret resolution failure (failed to fetch secret).
Implementation notes
- Review the source instructions and adapt inputs before running the workflow.
Try it, then tell us how it went
A copied prompt does not tell us whether the task worked. Report your result after checking the output. We collect your tool and outcome, never your files or prompt.