Soham Patankar
fbe1e9977f
All-platform progressive-disclosure skill split + generator (addresses #1106). Splits each platform's skill into a lean core (~615 lines, full default pipeline inline) + on-demand references/, generated from a single source via tools/skillgen with a CI/pre-commit drift gate. 13 hosts split, aider/devin stay monoliths. Also fixes the stale bare-path bugs across the previously hand-maintained variants and moves the always-on blocks into packaged markdown. Verified: all 5 generator guards pass, byte-verbatim load-bearing slices, lean cores self-sufficient on the default path across all 13 split hosts, references gated to non-default branches, description preserves the graphify-out-query-first clause. Supersedes #1119 (Claude-first subset). Known follow-up applied on top: harden _always_on() against a missing packaged file so a partial install can't brick the CLI.
1.7 KiB
Step B2 - Dispatch ALL subagents in a single message (Codex)
Codex platform: Uses
spawn_agent+wait_agent+close_agentinstead of the Agent tool. Requiresmulti_agent = trueunder[features]in~/.codex/config.toml. Ifspawn_agentis unavailable, tell the user to add that config and restart Codex.
Call spawn_agent once per chunk — ALL in the same response so they run in parallel. Build the message by wrapping the extraction prompt in task-delegation framing:
spawn_agent(agent_type="worker", message="Your task is to perform the following. Follow the instructions below exactly.\n\n<agent-instructions>\n[extraction prompt, with FILE_LIST, CHUNK_NUM, TOTAL_CHUNKS, DEEP_MODE substituted]\n</agent-instructions>\n\nExecute this now. Output ONLY the structured JSON response.")
After all agents are dispatched, collect results sequentially in memory:
result = wait_agent(handle); close_agent(handle) # repeat per handle
Parse each result as JSON. Accumulate nodes/edges/hyperedges across all results and write to graphify-out/.graphify_semantic_new.json. Codex collects in memory, so there are no per-chunk files on disk; the disk-based success checks in Step B3 do not apply — a chunk that returns invalid JSON is the failure signal instead.
Subagent prompt template:
See references/extraction-spec.md for the compact subagent prompt (rules, node-ID format, confidence rubric, hyperedge and vision rules, JSON schema). Load it only here, only when at least one chunk holds a doc, paper, or image; a pure-code corpus has skipped Part B and never reads it. Pass each agent that prompt verbatim with FILE_LIST, CHUNK_NUM, TOTAL_CHUNKS, and DEEP_MODE substituted, and have it return the JSON inline.