Skip to content

paradigmxyz/centaur / .agents/skills/centaur-observability

centaur-observability

Verified

Investigate Centaur runtime behavior with the VictoriaLogs (`vlogs`) and VictoriaMetrics (`vmetrics`) Code Mode connectors. Use when a user reports something broken or stalled, a workflow, alert, or channel post never populated, asks to check code for issues, or needs thread, execution, sandbox, tool-call, service-health, or deploy evidence.

FileKindSize
SKILL.mdskill md2.9 KB
---
name: centaur-observability
description: "Investigate Centaur runtime behavior with the VictoriaLogs (`vlogs`) and VictoriaMetrics (`vmetrics`) Code Mode connectors. Use when a user reports something broken or stalled, a workflow, alert, or channel post never populated, asks to check code for issues, or needs thread, execution, sandbox, tool-call, service-health, or deploy evidence."
---

# Centaur Observability

Investigate runtime evidence before proposing redesigns, simplifications, or config rewiring. Skip this skill if the sandbox prompt says observability access is disabled.

## Workflow

1. Identify the affected surface: thread key, execution id, service, tool, and rough time window.
2. Read the relevant code paths and workflow status alongside the logs.
3. Query the narrowest matching evidence below in Code Mode (`centaur-codemode exec --result-only`). Run independent queries together in one cell.
4. Report a root cause when established, or bounded hypotheses plus the evidence needed to confirm them. Offer redesign advice only after findings, or when the user asks to skip investigation.

For auth, credential, or proxy failures, use the `auth-failure-log-triage` skill.

## Logs

Functions on `tools.vlogs`:

```js
await tools.vlogs.errors({ start: "1h" });
await tools.vlogs.errors({ service: "api", start: "6h" });
await tools.vlogs.thread_logs({ thread_key: "slack:T1:C1:1234", start: "24h" });
await tools.vlogs.thread_trace({ thread_key: "slack:T1:C1:1234" });
await tools.vlogs.execution_timeline({ execution_id: "exe_123" });
await tools.vlogs.slow_requests({ threshold_ms: 3000 });
await tools.vlogs.tool_calls({ tool_name: "websearch", start: "24h" });
await tools.vlogs.tool_analytics({ start: "7d" });
await tools.vlogs.service_health({ start: "1h" });
await tools.vlogs.sandbox_activity({ start: "1h" });
await tools.vlogs.query({ query: "level:error AND event:tool_call_completed", limit: 20 });
```

For example, one cell for a stalled thread:

```js
const thread_key = "slack:T1:C1:1234";
const [trace, errors, tools_used] = await Promise.allSettled([
  tools.vlogs.thread_trace({ thread_key, start: "24h" }),
  tools.vlogs.errors({ thread_key, start: "24h" }),
  tools.vlogs.tool_usage_by_thread({ thread_key, start: "24h" }),
]);
return { trace, errors, tools_used };
```

## Metrics

```js
const exprs = [
  "centaur_deployment_info",                 // component, version, git_sha, image
  "centaur_last_deploy_timestamp_seconds",   // last deploy per component
  "centaur_overlay_revision_info",           // deployed overlay revisions
  "centaur_overlay_revision_scrape_success", // overlay scrape health
];
return await Promise.all(exprs.map((expr) => tools.vmetrics.query({ expr })));
```

Run `centaur-codemode load vlogs` or `centaur-codemode load vmetrics` for every function and its arguments. If a query fails or returns nothing, say so; do not substitute repo files, values files, or memory for live state.
writing-skillsobra/superpowersUse when creating new skills, editing existing skills, or verifying skills work before deploymentStars: 322kUpdated Claude CodeCodexVerifiedInstall
xlsxanthropics/skillsUse this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing .xlsx, .xlsm, .xltx, .csv, or .tsv file (e.g., adding columns, computing formulas, formatting, charting, cleaning messy data); create a new spreadsheet from scratch or from other data sources; or convert between tabular file formats. Trigger especially when the user references a spreadsheet file by name or path — even casually (like "the xlsx in my downloads") — and wants something done to it or produced from it. Also trigger for cleaning or restructuring messy tabular data files (malformed rows, misplaced headers, junk data) into proper spreadsheets. The deliverable must be a spreadsheet file. Do NOT trigger when the primary deliverable is a Word document, HTML report, standalone Python script, database pipeline, or Google Sheets API integration, even if tabular data is involved.Stars: 198kUpdated Claude CodeCodexVerifiedInstall
shipping-and-launchaddyosmani/agent-skillsPrepares production launches. Use when preparing to deploy to production, or when asking what needs to be in place before shipping. Use when you need a pre-launch checklist, when setting up monitoring, when planning a staged rollout, or when you need a rollback strategy.Stars: 119kUpdated Claude CodeCodexVerifiedInstall
childhoodthedotmack/claude-memBabysit a shipped release after merge, publication, or deployment. Verify published packages, production behavior, installed plugins, and late CI or user reports until the observation window is clean; use babysit for PR readiness before release.Stars: 118kUpdated Claude CodeCodexVerifiedInstall