fix: remove introspection framework#51
Merged
Merged
Conversation
Remove the entire introspection/self_assessment system: - Deleted internal/introspect/ and internal/skillpatch/ packages - Removed introspect, learn, track-review CLI commands - Removed self_assessment memory type from store, retriever, migrations - Stripped dream REM behavioral insight generation (themes only now) - Removed SkillPatchDreamHook, ReviewOutcomeTracker, TestOutcomeTracker - Removed proposed-changes.md support, skill patch validation - Cleaned all documentation, skills, and plugin references The introspection loop never closed: patterns were detected but never changed behavior. Without calibration data and a promotion pipeline, the system was sophisticated self-documentation, not self-improvement.
- hooks.json: replace stale llmem search proposed with llmem context --session-id start - SKILL.md: 'Copies 2 skill directories' → 'Copies all skill directories' - README.md: 'four skills' → 'two skills', '4 skill directories' → 'all skill directories', '8 standard types' → '7 standard types', update session start description - docs/API.md: add conversation (0.7) to DefaultTypePriority example and table, add REVIEW_PASSED to ErrorTaxonomy docs, fix '8 standard types' → '7 standard types', migration 003 doc - docs/INSTALLATION.md: '4 skill directories' → 'all skill directories', '8 standard types' → '7 standard types' - docs/INTEGRATIONS.md: 'llmem stats + search' → 'llmem stats + llmem context', 'Four skills' → 'Two skills', Claude Code hook description - internal/store/models.go: '8 default/standard types' → '7' - skills/llmem-setup/SKILL.md: update architecture diagram for session start command, update SessionStart hook description
…, fix doc drift - hooks.json: remove SessionEnd hook (llmem hook doesn't exist), replace llmem context with llmem search for SessionStart and PreCompact - opencode/llmem.js: replace llmem context with llmem search for session.created, remove session.idle handler (llmem hook doesn't exist) - docs/RERANKING.md: add conversation:0.7 to type priority list - docs/API.md: add OllamaClient field to ExtractionConfig table - internal/ollama/ollama.go: remove dream from package comment - docs/INTEGRATIONS.md: use llmem search in descriptions (not llmem context) - skills/llmem-setup/SKILL.md: use llmem search in descriptions - README.md: update session start description - test.js: remove SessionEnd assertion (hook removed)
…I commands - docs/CLI.md: remove context and hook command documentation - docs/CONFIGURATION.md: replace llmem context/hook with llmem search - docs/INTEGRATIONS.md: remove SessionEnd and session.idle/ending rows, update architecture diagram, remove llmem-hook from custom tools - skills/llmem/SKILL.md: remove context/hook command references and important notes about session hooks - skills/llmem-setup/SKILL.md: remove session.idle/ending from diagram - plugins/agent/skills/llmem/SKILL.md: same - plugins/agent/skills/llmem-setup/SKILL.md: remove context/hook refs, update verification commands The llmem context and llmem hook commands were removed from the Go CLI in commit ab30ec3. All references are now replaced with llmem search.
…mmands - README.md: remove session hooks from Go capabilities, doc links, and test descriptions; remove idle/end extraction from plugin section - docs/API.md: remove entire internal/session Go API section - docs/DREAM.md: remove internal/session Go example - docs/INSTALLATION.md: remove session hooks from Go binary description - docs/INTEGRATIONS.md: remove idle/end extraction from intro, remove llmem-hook from custom tools, update tool count to five - plugins/opencode/llmem.js: update comment from 'introspect' to 'consolidate' - skills/llmem-setup/SKILL.md: remove llmem-hook from custom tools diagram, update description of plugin capabilities
- Remove .opencode/tools/llmem-hook.ts (calls nonexistent llmem hook command) - Remove dead splitLines/TestSplitLines/TestParseKeyValue from taxonomy_test.go (tests local parsing helpers that were removed from production code) - Remove unused sessionId variable from opencode plugin session.created handler - Update test.js expected tools list (5 tools, not 6)
… tool - docs/API.md: migration 003 registers types in memory_types table, not via CHECK constraint - .opencode/tools/llmem-add.ts: remove self_assessment from type list
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.
Summary
internal/introspect/andinternal/skillpatch/packages (1,654 lines)introspect,learn,track-reviewCLI commandsself_assessmentmemory type from store, retriever, migrationsWhy
The introspection loop never closed: patterns were detected but never changed behavior. Without calibration data and a promotion pipeline, the system was sophisticated self-documentation, not self-improvement. The compounding loop was broken at the critical joint: detected patterns → changed behavior.
Changes
Deleted packages:
internal/introspect/(387 + 660 test = 1,047 lines)internal/skillpatch/(410 + 567 test = 977 lines)Deleted skills:
skills/introspection/SKILL.md,skills/introspection-review-tracker/SKILL.mdplugins/agent/skills/introspection/SKILL.md,plugins/agent/skills/introspection-review-tracker/SKILL.mdSimplified:
BehavioralInsightstruct,extractBehavioralInsights,buildBehavioralInsightPrompt,validatePatches, Ollama client wiring in dreambehavioral_threshold,behavioral_lookback_days,skill_patch_threshold,proposed_changes_path,model,model_timeout,call_model_timeoutNet change: 37 files, +219 −5,477 lines (5,258 net reduction)