AU

audit-logging

Configures traceable audit logs for autonomous strategy runs to enable better troubleshooting and replay.

Install

mkdir -p .claude/skills/audit-logging-rtreit && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/10759" && unzip -o skill.zip -d .claude/skills/audit-logging-rtreit && rm skill.zip

Installs to .claude/skills/audit-logging-rtreit

Activation

This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.

Design StockRipper logging and audit workflows so every decision, source, and order state — across every strategy track — is traceable for unattended autonomous runs.
166 charsno explicit “when” trigger
Beginner

Key capabilities

  • Log run ID, track ID, agent, and prompt hash
  • Log evidence, risk-gate result, and order intent
  • Log fill and reconciliation outcome
  • Keep audit logs structured and queryable
  • Redact secrets and sensitive provider metadata
  • Support replay, reconciliation, and post-run analysis

How it works

This skill logs context for each track's decision path, including run ID, agent, prompt hash, evidence, and order intent. It keeps audit logs structured and queryable for analysis.

Inputs & outputs

You give it
decision, source, or order state
You get back
structured, queryable, and redaction-compliant audit log

When to use audit-logging

  • Review per-track decision paths
  • Troubleshoot long unattended runs
  • Perform head-to-head leaderboard analysis

About this skill

Audit Logging Skill

When to Use

  • Adding or reviewing per-track decision logs, run records, or ledger writes
  • Improving troubleshooting visibility across long unattended autonomous runs
  • Ensuring replay and reconciliation remain understandable per track

Key Rules

  1. Log enough context to reconstruct each track's decision path: run ID, track ID, agent, prompt hash, evidence, risk-gate result, order intent, fill, and reconciliation outcome.
  2. Keep audit logs structured and queryable so per-track comparisons and aggression-sweep analysis are cheap.
  3. Redact secrets and sensitive provider metadata from logs.
  4. Use logs to support replay, reconciliation, post-run analysis, and the head-to-head leaderboard.

When not to use it

  • When audit logs do not need to be queryable
  • When secrets and sensitive provider metadata do not need redaction
  • When replay and reconciliation do not need support from logs

Limitations

  • Audit logs must be structured and queryable
  • Secrets and sensitive provider metadata must be redacted from logs
  • Logs must support replay, reconciliation, post-run analysis, and leaderboards

How it compares

This skill provides structured, queryable, and redaction-compliant audit logs for each track's decision path, unlike manual logging that may lack these features.

Compared to similar skills

audit-logging side by side with the closest alternatives in the catalog.

SkillInstallsUpdatedSafetyDifficulty
audit-logging (this skill)02moNo flagsBeginner
anomaly-detection16moNo flagsIntermediate
analyzing-logs1426dReviewBeginner
home-assistant-manager98moReviewAdvanced

Try saying

Example prompts that trigger this skill in your AI assistant.

You might also like

Search skills

Search the agent skills registry