Imported from speakASAP/logging-microservice (AGENTS.md). Install upstream with npx skills add speakASAP/logging-microservice. Copyright stays with the author.

Repository Agent Instructions

Shared rules live here:

• Codex profile: /home/ssf/.codex/AGENTS.md
• Cross-agent standard: /home/ssf/.ai-agent-standards/CROSS_AGENT_AUTOMATION_STANDARD.md
• Repository operations: AGENT_OPERATIONS.md

Read those first, then follow the repository-specific notes below and the current planning/status files.

Repository-Specific Notes

AGENTS.md — logging-microservice

Required Reading

Read BUSINESS.md, SYSTEM.md, docs/00_constitution/CONSTITUTION.md, docs/01_vision/VISION.md, TASKS.md, and STATE.json before any change.

Authority

BUSINESS.md is human-owned and immutable to AI agents. SYSTEM.md, CLAUDE.md, and IPS planning documents may be proposed/updated by agents subject to review.

Service-to-service authentication

For machine service identity, follow the sole canonical SERVICE_IDENTITY_CONSUMER_STANDARD.md. It is not reproduced here.

Intent Preservation System

This repository follows Vision -> Goal Impact -> System -> Feature -> Task -> Execution Plan -> Coding Prompt -> Code -> Validation per AGENT_OPERATIONS.md and the central intent-preservation-system standard.

Safety and Operations

Never delete log files, weaken retention/rotation, or expose secrets. Escalate to a human if /health fails or a rollout fails.

Project-Specific Rules

This service is a shared dependency of every other ecosystem service; treat any endpoint or schema change as a breaking-change candidate requiring ecosystem-wide review before deployment.

Knowledge Retrieval

Use docs-rag-microservice for bounded discovery when it is healthy, then verify deployment, security, database, integration and public-contract facts against the cited Git source. Git remains authoritative.

Authority and fallback rules: /home/ssf/Documents/Github/shared/docs/DOCUMENTATION_AUTHORITY.md.

Do not generate tokens in documentation or assume an unconfident/failed RAG response means that source documentation does not exist.

Automation Status

No active AI orchestration agents for this service. This service is a dependency of all other services in the ecosystem — changes to its API require ecosystem-wide review before deployment.

What Agents May Do

• Read all files for context
• Update TASKS.md and STATE.json
• Propose edits to SYSTEM.md and CLAUDE.md
• Run read-only kubectl commands: kubectl get, kubectl describe, kubectl logs
• Run scripts/deploy.sh (triggers docker build + kubectl rollout)

What Agents Must NOT Do

• Edit BUSINESS.md (human-owned)
• Push to remote git without human review

Deployment

# Build image, push to localhost:5000, trigger kubectl rollout
./scripts/deploy.sh

Operational Commands

# Live logs
kubectl logs -f deploy/logging-microservice -n statex-apps

# Pod status
kubectl get pods -n statex-apps -l app=logging-microservice

# ConfigMap contents
kubectl get configmap logging-microservice-config -n statex-apps -o yaml

# Verify Vault secret sync
kubectl get secret logging-microservice-secret -n statex-apps

# Health check
curl https://logging.alfares.cz/health

Escalation

If /health returns non-200 or a rollout fails, escalate to a human before retrying.

Required Final Report

Report files changed, validation evidence (adoption gate output), validation debt used or added, blockers as [MISSING: ...]/[UNKNOWN: ...], and the next concrete action.