How do I install the Audit skill?

Run npx skillmds@latest add ww-w-ai/audit in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Audit safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Audit?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Audit free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

ww-w-ai (@ww-w-ai) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Audit by ww-w-ai — Agent Skill

Name: Audit
Author: ww-w-ai

Audit

Query audit trail logs, decision traces, and session history for AI transparency. Provides searchable access to all bkit audit records via MCP tools. Use proactively when user wants to review past decisions, trace actions, or inspect logs. Triggers: audit, log, decision trace, audit trail, history, 감사, 로그, 결정 추적, 이력, 監査, ログ, 決定追跡, 履歴, 审计, 日志, 决策追踪, 历史, auditoría, registro, rastreo de decisiones, audit, journal, traçage des décisions, Audit, Protokoll, Entscheidungsverfolgung, audit, registro, tracciamento delle decisioni Do NOT use for: creating new audit entries (hooks do that automatically), modifying logs

ww-w-ai Updated Sep 17, 2026

File contents

Audit Skill

Query and review audit trail logs for transparency and decision tracing

Commands

Command	Description	Example
`/audit log`	Show recent audit entries	`/audit log`
`/audit search [query]`	Search audit logs by keyword	`/audit search "deploy"`
`/audit summary`	Summarize audit activity	`/audit summary --from 2026-04-01`

MCP Integration

This skill uses the bkit_audit_query MCP tool to retrieve audit data.

How to Execute

Show Recent Logs
- Call bkit_audit_query with action list
- Display results in a table format with timestamp, type, and summary
Search Logs
- Call bkit_audit_query with action search and the user's query
- Filter by date range if --from and --to are provided
- Filter by type if --type is provided (e.g., decision, action, error)
Generate Summary
- Call bkit_audit_query with action summary
- Aggregate entries by type and provide counts
- Highlight key decisions and their outcomes

Audit Entry Types

Type	Description
`decision`	Agent decision points with rationale
`action`	File modifications, tool invocations
`error`	Errors encountered during execution
`checkpoint`	Checkpoint creation/restoration events
`pdca`	PDCA phase transitions

Output Format

## Audit Log Report

### Period: {from} to {to}

| Timestamp | Type | Summary |
|-----------|------|---------|
| 2026-04-09 14:23 | decision | Selected gap-detector agent for verification |
| 2026-04-09 14:24 | action | Modified src/api/auth.js (lines 45-67) |

### Statistics
- Total entries: N
- Decisions: N | Actions: N | Errors: N

Data Location

Audit logs are stored in .bkit/audit/ directory:

.bkit/audit/sessions/ - Per-session audit trails
.bkit/audit/decisions/ - Decision trace records

Integration with PDCA

Audit is part of the Check phase:

Use /audit summary to review what happened during Do phase
Cross-reference with /pdca analyze results
Verify all decisions align with the Design document

---name: auditdescription: Query audit trail logs, decision traces, and session history for AI transparency. Provides searchable access to all bkit audit records via MCP tools. Use proactively when user wants to review past decisions, trace actions, or inspect logs. Triggers: audit, log, decision trace, audit trail, history, 감사, 로그, 결정 추적, 이력, 監査, ログ, 決定追跡, 履歴, 审计, 日志, 决策追踪, 历史, auditoría, registro, rastreo de decisiones, audit, journal, traçage des décisions, Audit, Protokoll, Entscheidungsverfolgung, audit, registro, tracciamento delle decisioni Do NOT use for: creating new audit entries (hooks do that automatically), modifying logs---# Audit Skill> Query and review audit trail logs for transparency and decision tracing## Commands| Command | Description | Example ||---------|-------------|---------|| `/audit log` | Show recent audit entries | `/audit log` || `/audit search [query]` | Search audit logs by keyword | `/audit search "deploy"` || `/audit summary` | Summarize audit activity | `/audit summary --from 2026-04-01` |## MCP IntegrationThis skill uses the `bkit_audit_query` MCP tool to retrieve audit data.### How to Execute1. **Show Recent Logs** - Call `bkit_audit_query` with action `list` - Display results in a table format with timestamp, type, and summary2. **Search Logs** - Call `bkit_audit_query` with action `search` and the user's query - Filter by date range if `--from` and `--to` are provided - Filter by type if `--type` is provided (e.g., `decision`, `action`, `error`)3. **Generate Summary** - Call `bkit_audit_query` with action `summary` - Aggregate entries by type and provide counts - Highlight key decisions and their outcomes## Audit Entry Types| Type | Description ||------|-------------|| `decision` | Agent decision points with rationale || `action` | File modifications, tool invocations || `error` | Errors encountered during execution || `checkpoint` | Checkpoint creation/restoration events || `pdca` | PDCA phase transitions |## Output Format```markdown## Audit Log Report### Period: {from} to {to}| Timestamp | Type | Summary ||-----------|------|---------|| 2026-04-09 14:23 | decision | Selected gap-detector agent for verification || 2026-04-09 14:24 | action | Modified src/api/auth.js (lines 45-67) |### Statistics- Total entries: N- Decisions: N | Actions: N | Errors: N```## Data LocationAudit logs are stored in `.bkit/audit/` directory:- `.bkit/audit/sessions/` - Per-session audit trails- `.bkit/audit/decisions/` - Decision trace records## Integration with PDCAAudit is part of the Check phase:1. Use `/audit summary` to review what happened during Do phase2. Cross-reference with `/pdca analyze` results3. Verify all decisions align with the Design document

Frequently asked questions