Open
Conversation
Replace raw chat transcript dump with a clean documentation pipeline: - Sanitize XML noise, interrupt markers, API errors from messages - Filter narration-only and empty messages, skip noise-only sessions - Merge consecutive same-role fragments and format as structured markdown - Add optional --reflect flag to generate LLM-powered session insights (learnings, key takeaway, team context, changes, discoveries) via Ollama with a customizable prompt template Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replace conversation trail output with LLM-synthesized knowledge docs: - Each session is sent to Ollama which produces structured articles with Summary, Changes, Decisions, Insights, and Context sections - Conversation truncated to 8k chars (keeping most recent) for model context - Reflect is on by default (--no-reflect to skip), falls back to cleaned conversation format when Ollama is unavailable - Generate .smriti/CLAUDE.md index for Claude Code auto-discovery - Scope .smriti/knowledge to smriti project only (no cross-project data) - Prompt template customizable at .smriti/prompts/share-reflect.md Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Document the full category system end-to-end: default category tree, auto-classification pipeline, manual tagging, custom categories, how tags filter across all commands, and category preservation through the share/sync roundtrip.
Address common new-user questions: knowledge capture, data access, AI agent querying, multi-project separation, issue tracker integrations, and feature preservation during changes.
Phase 1 of the structured memory pipeline: Claude Code JSONL transcripts are now parsed into 13 typed MessageBlock types (tool calls, file ops, git, commands, errors, thinking, etc.) and stored in 6 queryable sidecar tables alongside the existing FTS index. Backward-compatible — legacy parseClaudeJsonl() unchanged. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- `smriti context` generates a compact project summary (~200-300 tokens) from sidecar tables and injects it into .smriti/CLAUDE.md - `smriti compare` measures token/tool differences between sessions for A/B testing context effectiveness - Fix initSmriti to call initializeMemoryTables for fresh databases - README documents context as the north star with honest A/B test results and a guide for community testing Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Signed-off-by: Rahul Tripathi <rahul.psit.ec@gmail.com>
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
2 participants
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.
No description provided.