omega-stenographer-mcp
The server records and retrieves conversation history as structured notes, enabling search, summarization, and milestone tracking.
Ingest conversation turns (role, content) under a session ID.
Retrieve a live running notes document for the current session.
Get a compressed briefing plus top-k relevant fragments when context is large.
Mark a specific exchange as a critical milestone with a label.
Run FTS5 keyword searches across all ingested exchanges, with an optional result limit.
Compatible with CrewAI as an MCP client, allowing CrewAI agent sessions to be monitored and compressed for later retrieval.
Compatible with VS Code Copilot as an MCP client, allowing passive observation and context compression within the Copilot environment.
Compatible with LangChain as an MCP client, enabling integration with LangChain-based agents for session observation and memory.
Omega Stenographer MCP
Stenographer preserves conversation turns, extracts decisions and blockers, and builds searchable briefs without deleting the original exchanges.
Version 2.0 requires explicit session scope. Use the same session_id across ingest, brief, compaction, milestone and history calls. A global history search requires cross_session=true. Marking a milestone changes priority; it does not establish that a statement is true.
Install with python -m pip install .. Run omega-stenographer-mcp for explicit MCP ingestion, or python codex_capture.py to run the server with automatic visible-message capture from local Codex transcript files. The compatibility standalone entry point uses the same maintained implementation. No model API key is required for extraction or search.
Data defaults to ~/.omega-stenographer. OMEGA_CODEX_SESSIONS overrides the transcript directory; otherwise capture uses $CODEX_HOME/sessions or ~/.codex/sessions. The initial cutoff defaults to September 7, 2026 and can be set with OMEGA_CAPTURE_SINCE=YYYY-MM-DD. Older files are indexed at their current end, and new appended messages are subsequently captured. One OS lock permits a single observer across multiple MCP processes.
The observer excludes tool results, reasoning and configuration messages; persists source receipts and SQLite file offsets; retries partial lines; and delivers committed exchanges to Omega Brain through a durable local event bus. The two projects carry the same versioned runtime module so Stenographer also installs independently.
stenographer_capture_status reports progress and delivery errors. Start a helper with StenographerClient(session_id="example-task") and use ingest, get_brief, compact_guard, search or mark_milestone within that scope.
See current operating contract for migration, task isolation and backups. Older examples need explicit session IDs. Run python -m unittest discover -s tests -p "test_contracts.py" for standalone contract tests. SSWP software evidence uses the same task IDs and event bus.
Available Tools
6 toolsstenographer_capture_statusB
Report automatic capture progress and delivery errors.
Inputs: no named parameters.
stenographer_compact_guardC
Get compressed briefing + task-scoped relevant fragments. Call when context window pressure hits.
Inputs: session_id (required), query.
stenographer_get_briefC
Get the live running notes document.
Inputs: session_id (required).
stenographer_ingest_exchangeA
Ingest a conversation turn. Call after every significant user or assistant message. Automatically: extracts decisions/blockers via regex, runs NAFE failure-signature scan (heuristic flags do not establish truth), checks for cross-system CLAEG TERMINAL_SHUTDOWN events, and compresses every STENO_TURN_LIMIT turns into session-scoped tiered briefs. Pass trace_id from omega_preload_context to enable cross-system correlation across Omega Brain, SSWP, and Stenographer.
Inputs: role (required), content (required), session_id (required), trace_id.
stenographer_mark_milestoneC
Flag a critical decision for priority priority.
Inputs: label (required), session_id (required), exchange_id (required).
stenographer_query_historyB
FTS5 keyword search over all ingested exchanges.
Inputs: query (required), limit, session_id, cross_session.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 6 tools
Each tool has a distinct, non-overlapping purpose: querying history, checking status, ingesting exchanges, retrieving briefs, managing compaction, and marking milestones. No ambiguity in tool selection.
All tools share the consistent 'stenographer_' prefix and follow a clear verb_noun pattern (e.g., query_history, capture_status, ingest_exchange), making naming predictable and uniform.
With 6 tools, the set is well-scoped for a stenographer/note-taking domain, covering essential operations without being overly numerous or sparse.
The tool set covers the full lifecycle: ingestion, querying, status monitoring, brief retrieval, compaction, and milestone marking. No obvious missing operations for the stated purpose.
Maintenance
Related MCP Connectors
Persistent, verbatim, searchable memory for AI assistants — one memory across every MCP client.
MCP-native notes and memory for ChatGPT, Claude, and other AI tools.
OCR, transcription, file extraction, and image generation for AI agents via MCP.
MCP-native Trust Infrastructure for AI Agents. Persistent encrypted memory with Trust Quotient.
Related MCP Servers
- AlicenseAqualityCmaintenanceAI agent provenance, trust, and auditability layer. VERITAS multi-gate scoring, Cortex approval gates, S.E.A.L. hash-chain audit ledger, and semantic RAG with cryptographic provenance tracking for every decision an agent makes.275MIT
- AlicenseAqualityCmaintenanceSSWP MCP — Deterministic software attestation for AI-augmented development. Witness any repo through a 5-gate pipeline, adversarially probe dependencies, and produce self-verifying .sswp.json attestations. Fleet registry across 131 nodes with tamper-proof audit ledger and FTS5 search. Agent-native — one tool call from Hermes, Claude, or Cline.83MIT
- AlicenseNot gradedqualityBmaintenanceA local-first AI secretary that gathers your work context into private memory and enables AI agents to search and summarize it over MCP.10 npmMIT
- AlicenseAqualityDmaintenanceMCP server that stores AI agent reasoning traces permanently on 0G Storage, enabling immutable and verifiable decision logging.5MIT