Local Knowledge-First & Metadata Discovery

To prevent unnecessary exploratory terminal commands, token window exhaustion, and context loss during agentic development sessions, AI agents must adhere to the Local Knowledge-First Protocol. All project facts, architectural specifications, inventory mappings, and operational rules are indexed via OKF v0.1 YAML Frontmatter in .agents/skills/ and docs/.

🧭 The 5-Step Discovery Flow

AI agents must navigate the following protocol sequentially to answer questions, debug issues, or execute changes:

Step 1: Local Frontmatter & Metadata Search

Search local OKF headers using keywords on topics: or description: across docs/ and .agents/skills/ before running system terminal commands.

Step 2: Targeted File Viewing

Read specific file segments or targeted markdown documents locally instead of executing full-file dumps or remote probes.

Step 3: Temporal Verification Gate

Inspect the OKF timestamp metadata. If the local document appears contextually outdated, research external standards and present a comparison to the human operator.

Step 4: Human Verification & Knowledge Update

With human consensus, update the local OKF-compliant document to preserve spatial memory before executing infrastructure modifications.

Step 5: Terminal Execution Gate

Only execute remote terminal queries or run Ansible commands against the target inventory if applying planned updates or querying undocumented real-time state.

🔒 Mandatory Metadata Rules

1. Frontmatter Definition: Every Markdown file must begin with a YAML block starting on line 1 with --- and closing with ---.
2. Mandatory OKF Keys: Frontmatter must define okf_version, type, title, timestamp, and topics.
3. No Drift Policy: The set of directories under .agents/skills/ must exactly match the skills documented in AGENTS.md.
4. Standard Footer Compliance: All markdown documents must end with the standard ASIMP/DSOM footer to ensure consistent branding and legal coverage.