| name | kilocode-bringup-audit |
| description | Guide for tiered bring-up, smoke validation, and evidence capture during PMOVES service deployments on the KiloCode GLM lane. Use when starting services, running smoke tests, or validating CI/CD pipeline outputs. |
| keywords | ["bringup","smoke","audit","validation","deployment","health","kilocode","glm"] |
| version | 1.0.0 |
| category | PMOVES/KiloCode |
KiloCode Bringup Audit
Tiered service bring-up, smoke validation, and evidence capture for PMOVES deployments, optimized for the KiloCode GLM delivery lane.
Purpose
Execute tiered service bring-up validation, capture pass/fail evidence, and generate remediation queues for PMOVES infrastructure. Route complex failures through TensorZero coding_glm / coding_kilocode for root-cause analysis.
Capabilities
- ✨ Execute tiered Make targets for service bring-up
- 🔍 Run health endpoint validation across KiloCode-critical services
- 📊 Generate pass/fail matrix with evidence
- 🔧 Queue remediation items for failures
- 🤖 Route complex failures to GLM-5-Turbo for root cause analysis
Integration Points
- NATS Subject:
pmoves.health.check.v1
- Health Endpoints: TensorZero
:3030/health, Ollama :11434/api/tags, Cipher :8105/health
- Make Targets:
make -C pmoves kilo-health, make -C pmoves smoke, make -C pmoves smoke-gpu
- TensorZero Functions:
coding_glm, coding_kilocode
Workflow
Tier 1: KiloCode Health
make -C pmoves kilo-health
Tier 2: Core Smoke
make -C pmoves smoke
Tier 3: GPU Validation (Strict)
GPU_SMOKE_STRICT=true make -C pmoves smoke-gpu
Tier 4: Full Integration
make -C pmoves verify-all
Evidence Capture
For each validation step, capture:
- Timestamp - ISO 8601 UTC
- Service - Name from
services-catalog.md
- Endpoint - Health path checked
- Status - HTTP status code or healthy/unhealthy
- Response - JSON body (truncated if large)
- Duration - Milliseconds
Remediation Queue
Failed checks generate remediation items with:
- Service name
- Expected vs actual status
- Logs snippet (
docker compose logs -f <service>)
- Suggested action
- Route to
coding_glm if failure pattern is novel
Example Usage
User: "Run bringup audit for workstation_5090 profile"
Agent:
1. Loads skill
2. Executes make -C pmoves kilo-health
3. Executes make -C pmoves smoke
4. Captures evidence to /tmp/bringup_audit_<iso>.json
5. Generates pass/fail matrix
6. Routes failures to TensorZero coding_glm for RCA
Trigger Phrases
- "run kilocode bringup audit"
- "smoke test services"
- "validate deployment"
- "service health check"
- "tiered bringup"
- "kilo-health"