Contour (知界) — Fine-grained cognitive boundary tracking for Claude Code.
Scan the current session for cognition/thinking/preference signals and write to extract-buffer.md. Run in the target session (current or --resume'd). Manual invocation only.
Cold-start initialization: generate Core Profile, Domain State, and Extract Buffer files, inject CLAUDE.md monitoring instructions. Run once per user, or to reset.
Read extract-buffer.md, distribute signals to Domain State, surface thinking patterns and core candidates in report, then clear buffer. Must run in a NEW dedicated session.
Remove Contour monitoring injection from CLAUDE.md, optionally delete data files. Safe: never deletes non-Contour files.
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
npx claudepluginhub alexu0317-father/contour --plugin contourDiagnose local proxy, Clash/Mihomo, routing rule, DNS, TUN, and app bypass problems with permissioned local probes and narrow reversible fixes.
AI mentor agent for vibe coders - analyzes Claude Code conversations with 6-axis skill evaluation (DECOMP/VERIFY/ORCH/FAIL/CTX/META), 7-level system (L1~L7, 0.5 increments), 4 workspace types (Builder/Explorer/Designer/Operator), and longitudinal growth tracking.
Skill memory layer for Claude Code — auto-capture, learn, and reuse skills from Acontext
I (Claude) act as you - Build knowledge base from your notes.
Self-Evolving AI Coding Infrastructure — Generate, Curate, and Enhance Reusable Wisdom
Agent skills that package evidence-backed pedagogical methodologies (explain-and-check, quiz-me, connect-to-what-you-know, ask-me-questions, learn-by-doing, linked-notes, flashcards) as workflows applied to code. The anti-cognitive-surrender layer: closes the comprehension gap that opens when an LLM has done the work on the human's behalf.
Ultra-compressed communication mode. Cuts ~75% of tokens while keeping full technical accuracy by speaking like a caveman.