Menu
Detects anti-patterns in agent workflows and computes health scores
Installer avec Codex ou Claude Copiez ce prompt, collez-le dans Codex, Claude ou un autre assistant, puis laissez-le vérifier la page du skill et l'installer pour vous.
Basé sur la classification professionnelle SOC
| name | workflow-insights |
| description | Detects anti-patterns in agent workflows and computes health scores |
| type | semantic |
| user-invocable | true |
| semantic | {"skillVersion":1,"permissions":{"graph":["read"],"memory":["recall"]},"queries":[{"predicate":"workflow:analysis","scope":"agent"}]} |
Detect anti-patterns in agent workflows and compute health scores based on 7 weighted pattern detectors.
Use when analyzing agent execution traces, debugging sluggish workflows, or auditing multi-step agent pipelines for inefficiency. The skill evaluates delegation chains, error patterns, resource usage, and tool utilization to produce a 0-100 health score.
| # | Pattern | Weight | Description |
|---|---|---|---|
| 1 | Repeated Failure | 20 | Same error appearing 3+ times suggests retry without fix |
| 2 | Long Chain | 15 | Delegation chain exceeding 5 steps suggests over-decomposition |
| 3 | Unused Tool | 5 | Tool registered but never called suggests bloated configuration |
| 4 | Redundant Query | 10 | Same query pattern appearing multiple times suggests missing caching |
| 5 | Timeout Pattern | 20 | Multiple timeouts in sequence suggests resource contention |
| 6 | Resource Waste | 15 | Large context sent to simple tasks suggests poor task routing |
| 7 | Error Cascade | 15 | Error in one step causing errors in 3+ downstream steps suggests missing error boundaries |
skill_memory_context using predicate workflow:analysisConnect the user's Akashi second brain (local, private, verifiable) over MCP and answer from cited vault passages with signed provenance. Query it BEFORE answering anything about the user's knowledge, projects, decisions, or conventions. Triggers - akashi, vault, my notes, second brain, remember this, what did I decide.
Use CodexBar CLI local cost usage to summarize per-model usage for Codex or Claude, including the current (most recent) model or a full model breakdown. Trigger when asked for model-level usage/cost data from codexbar, or when you need a scriptable per-model summary from codexbar cost JSON.
Create or update AgentSkills. Use when designing, structuring, or packaging skills with scripts, references, and assets.
Monitor API health and assert endpoint status with ZK proofs
Semantic code review with provable assertions about code quality
Audit project dependencies with verifiable security assertions