{ "schemaVersion": "1.0", "item": { "slug": "agent-audit", "name": "Agent Audit", "source": "tencent", "type": "skill", "category": "安全合规", "sourceUrl": "https://clawhub.ai/sharbelayy/agent-audit", "canonicalUrl": "https://clawhub.ai/sharbelayy/agent-audit", "targetPlatform": "OpenClaw" }, "install": { "downloadMode": "redirect", "downloadUrl": "/downloads/agent-audit", "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=agent-audit", "sourcePlatform": "tencent", "targetPlatform": "OpenClaw", "installMethod": "Manual import", "extraction": "Extract archive", "prerequisites": [ "OpenClaw" ], "packageFormat": "ZIP package", "includedAssets": [ "SKILL.md", "references/model-pricing.md", "references/task-classification.md", "scripts/audit.py" ], "primaryDoc": "SKILL.md", "quickSetup": [ "Download the package from Yavira.", "Extract the archive and review SKILL.md first.", "Import or place the package into your OpenClaw setup." ], "agentAssist": { "summary": "Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.", "steps": [ "Download the package from Yavira.", "Extract it into a folder your agent can access.", "Paste one of the prompts below and point your agent at the extracted folder." ], "prompts": [ { "label": "New install", "body": "I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Tell me what you changed and call out any manual steps you could not complete." }, { "label": "Upgrade existing", "body": "I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Summarize what changed and any follow-up checks I should run." } ] }, "sourceHealth": { "source": "tencent", "slug": "agent-audit", "status": "healthy", "reason": "direct_download_ok", "recommendedAction": "download", "checkedAt": "2026-04-24T21:03:15.019Z", "expiresAt": "2026-05-01T21:03:15.019Z", "httpStatus": 200, "finalUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=agent-audit", "contentType": "application/zip", "probeMethod": "head", "details": { "probeUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=agent-audit", "contentDisposition": "attachment; filename=\"agent-audit-1.0.0.zip\"", "redirectLocation": null, "bodySnippet": null, "slug": "agent-audit" }, "scope": "item", "summary": "Item download looks usable.", "detail": "Yavira can redirect you to the upstream package for this item.", "primaryActionLabel": "Download for OpenClaw", "primaryActionHref": "/downloads/agent-audit" }, "validation": { "installChecklist": [ "Use the Yavira download entry.", "Review SKILL.md after the package is downloaded.", "Confirm the extracted package contains the expected setup assets." ], "postInstallChecks": [ "Confirm the extracted package includes the expected docs or setup files.", "Validate the skill or prompts are available in your target agent workspace.", "Capture any manual follow-up steps the agent could not complete." ] }, "downloadPageUrl": "https://openagent3.xyz/downloads/agent-audit", "agentPageUrl": "https://openagent3.xyz/skills/agent-audit/agent", "manifestUrl": "https://openagent3.xyz/skills/agent-audit/agent.json", "briefUrl": "https://openagent3.xyz/skills/agent-audit/agent.md" }, "agentAssist": { "summary": "Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.", "steps": [ "Download the package from Yavira.", "Extract it into a folder your agent can access.", "Paste one of the prompts below and point your agent at the extracted folder." ], "prompts": [ { "label": "New install", "body": "I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Tell me what you changed and call out any manual steps you could not complete." }, { "label": "Upgrade existing", "body": "I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Summarize what changed and any follow-up checks I should run." } ] }, "documentation": { "source": "clawhub", "primaryDoc": "SKILL.md", "sections": [ { "title": "Agent Audit", "body": "Scan your entire OpenClaw setup and get actionable cost/performance recommendations." }, { "title": "What This Skill Does", "body": "Scans config — reads OpenClaw config to map models to agents/tasks\nAnalyzes cron history — checks every cron job's model, token usage, runtime, success rate\nClassifies tasks — determines complexity level of each task\nCalculates costs — per agent, per cron, per task type using provider pricing\nRecommends changes — with confidence levels and risk warnings\nGenerates report — markdown report with specific savings estimates" }, { "title": "Running the Audit", "body": "python3 {baseDir}/scripts/audit.py\n\nOptions:\n\npython3 {baseDir}/scripts/audit.py --format markdown # Full report (default)\npython3 {baseDir}/scripts/audit.py --format summary # Quick summary only\npython3 {baseDir}/scripts/audit.py --dry-run # Show what would be analyzed\npython3 {baseDir}/scripts/audit.py --output /path/to/report.md # Save to file" }, { "title": "Phase 1: Discovery", "body": "Read OpenClaw config (~/.openclaw/openclaw.json or similar)\nList all cron jobs and their configurations\nList all agents and their default models\nDetect provider (Anthropic, OpenAI, Google, xAI) from model names" }, { "title": "Phase 2: History Analysis", "body": "Pull cron job run history (last 7 days by default)\nCalculate per-job: avg tokens, avg runtime, success rate, model used\nPull session history where available\nCalculate total token spend by model tier" }, { "title": "Phase 3: Task Classification", "body": "Classify each task into complexity tiers:\n\nTierExamplesRecommended ModelsSimpleHealth checks, status reports, reminders, notificationsCheapest tier (Haiku, GPT-4o-mini, Flash, Grok-mini)MediumContent drafts, research, summarization, data analysisMid tier (Sonnet, GPT-4o, Pro, Grok)ComplexCoding, architecture, security review, nuanced writingTop tier (Opus, GPT-4.5, Ultra, Grok-2)\n\nClassification signals:\n\nSimple: Short output (<500 tokens), low thinking requirement, repetitive pattern, status/health tasks\nMedium: Medium output, some reasoning needed, creative but templated, research tasks\nComplex: Long output, multi-step reasoning, code generation, security-critical, tasks that previously failed on weaker models" }, { "title": "Phase 4: Recommendations", "body": "For each task where the model tier doesn't match complexity:\n\n⚠️ RECOMMENDATION: Downgrade \"Knox Bot Health Check\" from opus to haiku\n Current: anthropic/claude-opus-4 ($15/M input, $75/M output)\n Suggested: anthropic/claude-haiku ($0.25/M input, $1.25/M output)\n Reason: Simple status check averaging 300 output tokens\n Estimated savings: $X.XX/month\n Risk: LOW — task is simple pattern matching\n Confidence: HIGH" }, { "title": "Safety Rules — NEVER Recommend Downgrading:", "body": "Coding/development tasks\nSecurity reviews or audits\nTasks that have previously failed on weaker models\nTasks where the user explicitly chose a higher model\nComplex multi-step reasoning tasks\nAnything the user flagged as critical" }, { "title": "Phase 5: Report Generation", "body": "Output a clean markdown report with:\n\nOverview — total agents, crons, monthly spend estimate\nPer-agent breakdown — model, usage, cost\nPer-cron breakdown — model, frequency, avg tokens, cost\nRecommendations — sorted by savings potential\nTotal potential savings — monthly estimate\nOne-liner config changes — exact model strings to swap" }, { "title": "Model Pricing Reference", "body": "See references/model-pricing.md for current pricing across all providers.\nUpdate this file when prices change." }, { "title": "Task Classification Details", "body": "See references/task-classification.md for detailed heuristics\non how tasks are classified into complexity tiers." }, { "title": "Important Notes", "body": "This skill is read-only — it never changes your config automatically\nAll recommendations include risk levels and confidence scores\nWhen unsure about a task's complexity, it defaults to keeping the current model\nThe audit should be re-run periodically (monthly) as usage patterns change\nToken counts are estimates based on cron history — actual costs depend on your provider's billing" } ], "body": "Agent Audit\n\nScan your entire OpenClaw setup and get actionable cost/performance recommendations.\n\nWhat This Skill Does\nScans config — reads OpenClaw config to map models to agents/tasks\nAnalyzes cron history — checks every cron job's model, token usage, runtime, success rate\nClassifies tasks — determines complexity level of each task\nCalculates costs — per agent, per cron, per task type using provider pricing\nRecommends changes — with confidence levels and risk warnings\nGenerates report — markdown report with specific savings estimates\nRunning the Audit\npython3 {baseDir}/scripts/audit.py\n\n\nOptions:\n\npython3 {baseDir}/scripts/audit.py --format markdown # Full report (default)\npython3 {baseDir}/scripts/audit.py --format summary # Quick summary only\npython3 {baseDir}/scripts/audit.py --dry-run # Show what would be analyzed\npython3 {baseDir}/scripts/audit.py --output /path/to/report.md # Save to file\n\nHow It Works\nPhase 1: Discovery\nRead OpenClaw config (~/.openclaw/openclaw.json or similar)\nList all cron jobs and their configurations\nList all agents and their default models\nDetect provider (Anthropic, OpenAI, Google, xAI) from model names\nPhase 2: History Analysis\nPull cron job run history (last 7 days by default)\nCalculate per-job: avg tokens, avg runtime, success rate, model used\nPull session history where available\nCalculate total token spend by model tier\nPhase 3: Task Classification\n\nClassify each task into complexity tiers:\n\nTier\tExamples\tRecommended Models\nSimple\tHealth checks, status reports, reminders, notifications\tCheapest tier (Haiku, GPT-4o-mini, Flash, Grok-mini)\nMedium\tContent drafts, research, summarization, data analysis\tMid tier (Sonnet, GPT-4o, Pro, Grok)\nComplex\tCoding, architecture, security review, nuanced writing\tTop tier (Opus, GPT-4.5, Ultra, Grok-2)\n\nClassification signals:\n\nSimple: Short output (<500 tokens), low thinking requirement, repetitive pattern, status/health tasks\nMedium: Medium output, some reasoning needed, creative but templated, research tasks\nComplex: Long output, multi-step reasoning, code generation, security-critical, tasks that previously failed on weaker models\nPhase 4: Recommendations\n\nFor each task where the model tier doesn't match complexity:\n\n⚠️ RECOMMENDATION: Downgrade \"Knox Bot Health Check\" from opus to haiku\n Current: anthropic/claude-opus-4 ($15/M input, $75/M output)\n Suggested: anthropic/claude-haiku ($0.25/M input, $1.25/M output)\n Reason: Simple status check averaging 300 output tokens\n Estimated savings: $X.XX/month\n Risk: LOW — task is simple pattern matching\n Confidence: HIGH\n\nSafety Rules — NEVER Recommend Downgrading:\nCoding/development tasks\nSecurity reviews or audits\nTasks that have previously failed on weaker models\nTasks where the user explicitly chose a higher model\nComplex multi-step reasoning tasks\nAnything the user flagged as critical\nPhase 5: Report Generation\n\nOutput a clean markdown report with:\n\nOverview — total agents, crons, monthly spend estimate\nPer-agent breakdown — model, usage, cost\nPer-cron breakdown — model, frequency, avg tokens, cost\nRecommendations — sorted by savings potential\nTotal potential savings — monthly estimate\nOne-liner config changes — exact model strings to swap\nModel Pricing Reference\n\nSee references/model-pricing.md for current pricing across all providers. Update this file when prices change.\n\nTask Classification Details\n\nSee references/task-classification.md for detailed heuristics on how tasks are classified into complexity tiers.\n\nImportant Notes\nThis skill is read-only — it never changes your config automatically\nAll recommendations include risk levels and confidence scores\nWhen unsure about a task's complexity, it defaults to keeping the current model\nThe audit should be re-run periodically (monthly) as usage patterns change\nToken counts are estimates based on cron history — actual costs depend on your provider's billing" }, "trust": { "sourceLabel": "tencent", "provenanceUrl": "https://clawhub.ai/sharbelayy/agent-audit", "publisherUrl": "https://clawhub.ai/sharbelayy/agent-audit", "owner": "sharbelayy", "version": "1.0.0", "license": null, "verificationStatus": "Indexed source record" }, "links": { "detailUrl": "https://openagent3.xyz/skills/agent-audit", "downloadUrl": "https://openagent3.xyz/downloads/agent-audit", "agentUrl": "https://openagent3.xyz/skills/agent-audit/agent", "manifestUrl": "https://openagent3.xyz/skills/agent-audit/agent.json", "briefUrl": "https://openagent3.xyz/skills/agent-audit/agent.md" } }