action-validator
FreePre-action boundary checking — validates agent tool calls against declared capabilities and task contracts
The English summary is being prepared. Showing the original description.
baekenough/second-brain15Updated Oct 8, 2026
FreePublished on GitHub
model-escalation
Advisory model escalation based on task outcome tracking
The English summary is being prepared. Showing the original description.
See how to installFiles included (1)
- SKILL.md1.7 KB
SKILL.md (original)
Review the instructions the agent will receive before you install.
Model Escalation Skill
Tracks task outcomes and advises model upgrades when failures are detected. Advisory-only — the orchestrator makes the final decision (R010).
Escalation Path
haiku → sonnet → opus
Trigger Conditions
| Condition | Action |
|---|---|
| 2+ failures with same model for same agent type | Advise escalation |
| 3+ consecutive failures across any agent type | Advise global escalation |
| Sustained success after escalation | Advise de-escalation |
Thresholds
| Parameter | Default | Description |
|---|---|---|
failure_threshold | 2 | Failures before escalation advisory |
consecutive_threshold | 3 | Consecutive failures for global advisory |
cooldown_tasks | 5 | Successes before de-escalation advisory |
Cost Guard
- Advisory includes estimated cost multiplier
- De-escalation suggested after sustained success at higher tier
- Cost tracked per session via PPID-scoped temp file
Architecture
PostToolUse (Task) → task-outcome-recorder.sh
Records: agent_type, model, success/failure, error_summary
Storage: /tmp/.claude-task-outcomes-$PPID (JSON lines, max 50)
PreToolUse (Task) → model-escalation-advisor.sh
Reads outcomes → counts failures → advises escalation via stderr
Advisory only — never blocks, never modifies tool input
Advisory Format
--- [Model Escalation Advisory] ---
Agent type: {agent_type}
Current model: {current_model}
Recent failures: {count}/{threshold}
⚡ Recommended: Escalate to {next_model}
Cost impact: {multiplier} per task
---
Reviews
No reviews yet. Tell others what you thought.
Comments
No login needed. Share how it worked for you, ask a question, or note where it did not run.
No comments yet. Be the first to write one.
More skills from this repository
adaptive-harness
FreeAuto-detect project context and optimize harness — deactivate unused agents/skills, suggest missing experts, generate project profile
The English summary is being prepared. Showing the original description.
baekenough/second-brain15Updated Oct 8, 2026
Adversarial code review using attacker mindset — trust boundary, attack surface, business logic, and defense evaluation
The English summary is being prepared. Showing the original description.
baekenough/second-brain15Updated Oct 8, 2026
Apache Airflow best practices for DAG authoring, testing, and production deployment
The English summary is being prepared. Showing the original description.
baekenough/second-brain15Updated Oct 8, 2026
Alembic migration patterns for naming conventions, safety checks, expand-contract, env.py configuration, and CI integration
The English summary is being prepared. Showing the original description.
baekenough/second-brain15Updated Oct 8, 2026
ambiguity-gate
FreePre-routing ambiguity analysis — scores request clarity and asks clarifying questions when needed (inspired by ouroboros)
The English summary is being prepared. Showing the original description.
baekenough/second-brain15Updated Oct 8, 2026