Return the Jambavan operating protocol: recall memory, index/watch code, retrieve context before edits, patch surgically, run checks, and store durable decisions. Call once at the start of every host session. Includes recent project memories by default.

Build or refresh the Jambavan codebase index. Parses source files with a tree-sitter AST extractor and stores symbols in a local SQLite database. Run once per project, then incrementally on file changes. Returns indexing statistics (files processed, symbols extracted, duration).

Search the Jambavan index for code symbols and snippets most relevant to a query. Returns a ranked, token-budgeted context block of matching functions, classes, and types. Inject this block into your prompt to give the model precise, token-efficient codebase knowledge. Much cheaper than reading whole files — only the relevant symbol bodies are returned.

Control the live file watcher that keeps the index in sync as you edit code. Actions: "start" — begin watching (index must exist first); "stop" — stop watching; "status" — show watcher state. While running, every file save triggers an incremental re-index of just that file (no full rescan). Use "start" after jambavan_index, then forget about it — the index stays fresh automatically.

Activate Vibhishana Niti in the current session: efficient, truthful senior-dev rules. Returns a concise ruleset: YAGNI-first, stdlib before dependencies, shortest working diff, root-cause fixes. Levels: "lite" — build what's asked, name the leaner option; "full" — the default ladder; "ultra" — deletion extremist. Inject the returned text into your system prompt or next message.

Harvest every rin marker (lines marked with a ceiling and upgrade path) from the project into a debt ledger. Groups findings by file and flags any marker with no upgrade trigger — those are the ones that silently rot. Read-only. Call before a release or refactor sprint.

Build a lightweight knowledge graph from the current code index and return hub nodes plus edge confidence notes. Call jambavan_index first. Edges are structural contains plus capped inferred symbol-name mentions. Defaults to the first 5000 indexed symbols; higher symbol_limit costs more and may still omit very common inferred mention names.

Query the current knowledge graph: find matching nodes and BFS neighbors within a token budget. Call jambavan_index first. Uses extracted call/import edges where available plus inferred mentions.

Find the shortest path between two graph nodes/symbols using BFS over extracted and inferred edges. Call jambavan_index first.

Show which languages are backed by tree-sitter (full AST) vs regex fallback, plus current index statistics. Useful for verifying the parser setup after a fresh install.

Persist a memory as an Open Knowledge Format (OKF) concept document. Memories are markdown files with YAML frontmatter — human-readable, git-diffable, portable. Each memory has a title, body (markdown), optional tags, and a scope (e.g. project name). Memories with the same title in the same scope are overwritten (idempotent). Returns the OKF concept ID (scope/slug) of the stored document.

Search stored memories using BM25 full-text ranking. Searches across title, description, tags, and body content. Returns ranked results with score, title, scope, and a body preview. Optional: filter to a specific scope.

Recall all memories for a scope — the session wake-up operation. Returns the full content of every memory in the scope, newest first. Use at the start of a new session to restore context about a project or topic. Omit scope to retrieve all memories across all scopes.

Mine durable facts, decisions, TODOs, and constraints from a pasted session transcript or log. Stores each extracted item as an OKF memory in the requested scope. This is deterministic text mining — no summarization or external service.

Mark a memory as invalidated/superseded without deleting the OKF document. Use when a durable fact changed and you want temporal history instead of silent overwrite.

Delete a memory by its OKF concept ID (scope/slug), or delete all memories in a scope. Provide id to delete a single memory. Provide scope with delete_scope: true to wipe all memories in that scope. Returns confirmation with the count of deleted memories.

Show memory bundle statistics: total memory count and breakdown by scope. Read-only. Use to understand what is stored before a search or recall.

Read the contents of a project file. Paths are confined to the project root unless JAMBAVAN_ALLOW_OUTSIDE_ROOT=1.

Search project files using ripgrep, with grep fallback. Fast regex search. Paths are confined to the project root unless JAMBAVAN_ALLOW_OUTSIDE_ROOT=1.

List project files in a directory. Paths are confined to the project root unless JAMBAVAN_ALLOW_OUTSIDE_ROOT=1.