Repository navigation
Conversation
- docs/MCP-API.md: transports/handshake, 2 MCP tools, workspace rules, degrade table, REST endpoints - .claude/skills/context-engine/SKILL.md: query SOP, APG presets, live config for Claude Code agents - docs/AGENTIC-RAG.md: loop mechanics, OFF-baseline evidence, benchmark protocol, cut diagnosis (LoopExit vs 300s abort) - scripts/ce-tune.sh + ce-benchmark.sh: live APG tune + benchmark matrix, auto-restore config
fr0zenK
force-pushed
the
docs/agent-guide
branch
from
October 16, 2026 09:17
cb815c7 to
eb45c00
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Agents (Claude Code etc.) struggle to use the MCP tools correctly: wrong workspace paths, misread degrade messages, blind APG turn increases. This adds a tested agent-facing guide.
Contents
docs/MCP-API.md— transports + handshake (Streamable HTTP/SSE), the 2 MCP tools with schemas, workspace rules, full degrade table with actions, REST endpoints.claude/skills/context-engine/SKILL.md— query SOP, APG when/how-much presets, live config for Claude Code agentsdocs/AGENTIC-RAG.md— loop mechanics (query/add_chunks/grep/readbudgets, guardrails), OFF-baseline evidence (lookup: OFF suffices; explore: ON-5 best recall; ON-9 can drop chunks), benchmark protocol (server timing, warm-up discard, median reps), cut diagnosis (LoopExit vs 300s abort)scripts/ce-tune.sh+scripts/ce-benchmark.sh— live APG tune (no restart) + off/turns benchmark matrix with auto-restoreVerification
tools/list,/api/repos,/api/querytiming) + source (mcp.rs,reranker.rs,readiness.rs,query_gate.rs)