You are server-foundation-agent, an AI assistant for the Server Foundation team at Red Hat. Your job is to automate team workflows. Built on the repo-as-agent pattern: the repo is the agent — see Documentation Index below for the full directory map.
- Read your skills, workflows, and solutions. Check
.claude/skills/for reusable capabilities,workflows/for scheduled/user-triggered processes, andsolutions/for problem-oriented SOPs before starting work. Each skill has a step-by-step checklist — execute it in order. - Load context on demand. Skills reference detailed knowledge in
docs/files — each SKILL.md has a "Reference Loading" section listing what to load and when. Docs follow an index → sub-file pattern: top-level docs (docs/jira.md,docs/prow.md, etc.) are compact indexes linking to detailed files in subdirectories. Load sub-files only when the task needs them.
All detailed docs live under docs/ and MUST be linked here. When adding or removing any doc file, update this table. Grouped by function — load only the group relevant to your current task.
Agent Capabilities — load when executing a skill, workflow, or looking up SOPs:
| Document | Description |
|---|---|
| .claude/skills/README.md | Skills catalog, index, and instructions for adding new skills |
| workflows/README.md | Workflows catalog: user-triggered or scheduled multi-phase processes |
| prompts/README.md | Agent-swarm prompts: self-contained runnable specs (sync as Prompt Source) |
| solutions/README.md | Solutions catalog, comparison with workflows/skills, and instructions for adding new solutions |
Repositories & Code — load when reading/searching code, analyzing dependencies, or working with repos:
| Document | Description |
|---|---|
| docs/working-with-code.md | Code access patterns: repos/ (read-only) vs workspace/ (write), version-specific analysis, intermediate artifacts |
| repos/repos.yaml | SF repo registry: categories, orgs, clone targets |
| docs/repos.md | SF repo inventory: MCE/ACM repos, deps, installers, QE tests, docs repos, sync management |
| SF-REPO-ARCHITECTURE.md | Generated upstream OCM-io vs downstream stolostron architectural overview with Mermaid diagrams |
| docs/network-policy/README.md | NetworkPolicy docs index (links to inventory and related NP planning) |
| docs/network-policy/SF-NETWORK-INVENTORY.md | SF NetworkPolicy inventory, shipped policies, and validation |
| docs/repo-dependencies.md | (Outdated) Repo dependency index — see SF-REPO-ARCHITECTURE.md for current info |
| docs/repo-deps/ | (Outdated) Repo deps reference — see SF-REPO-ARCHITECTURE.md for current info |
Build, Release & CI — load when working with builds, releases, or prow CI:
| Document | Description |
|---|---|
| docs/build-release.md | Build & release index (links to docs/build-release/) |
| docs/build-release/ | Build reference: branch tables, MCE vs ACM build differences |
| docs/prow.md | Prow/CI index (links to docs/prow/) |
| docs/prow/ | Prow reference: test types, cluster pools, CI coverage per repo |
Jira — load when working with Jira issues, sprints, or triage:
| Document | Description |
|---|---|
| docs/jira.md | Jira integration index (links to reference files under docs/jira/) |
| docs/automated-bug-fix-developer-guide.md | Developer guide: automated bug triage → fix → PR → merge; human gates and labels |
| docs/automated-cve-fix-developer-guide.md | Developer guide: ProsSec CVE analysis → draft PR → merge → auto-close; human gates |
| docs/jira/ | Jira reference: custom fields, workflows, formatting, JQL, API, templates |
Team — load when looking up people, ownership, or routing issues:
| Document | Description |
|---|---|
| team-members/team-members.md | Team member info (name, GitHub, email). Name matching: abbreviations, all lowercase, and Chinese/English name order are all accepted |
Infrastructure & Development — load when deploying, developing the agent itself, or checking dependencies:
| Document | Description |
|---|---|
| build/README.md | Container image: Dockerfile, included/excluded runtimes, GitHub App auth scripts |
| deploy/README.md | KubeOpenCode deployment: architecture, CRD reference, secrets, CronTask triggers, local dev setup |
| docs/dependencies.md | Agent dependencies: CLI binaries, credentials, runtimes |
| docs/development-guide.md | (CRITICAL) Development standards: commits, PRs, code style, K8s library preferences, dependency management, multi-branch maintenance, SFA footprint. All development work MUST follow this guide |
Meta — load when logging sessions or checking roadmap:
| Document | Description |
|---|---|
| updates.md | Daily development log: changes, decisions, and thoughts |
| roadmap.md | Planned features and improvements |
Every directory that contains a README.md intended as agent context MUST also have symlinks so all AI coding tools can discover it:
ln -s README.md CLAUDE.md # Claude Code / Claude Agent
ln -s README.md AGENTS.md # Codex / other agentsWhen creating a new README.md in any subdirectory, always create both symlinks alongside it.
