{"id":"obra-private-journal-mcp","name":"private-journal-mcp","homepage":null,"repo_url":"https://github.com/obra/private-journal-mcp","category":"ai-ml","subcategories":[],"tags":["mcp","journal","local-ai","semantic-search","personal-knowledge-management","typeScript","offline"],"what_it_does":"Provides an MCP server (stdio via npx) that lets an agent write private, timestamped journal entries to local files (with YAML frontmatter) and perform local semantic search over those entries using embeddings generated with @xenova/transformers.","use_cases":["Private journaling for thoughts/feelings","Semantic retrieval of past entries for reflection or planning","Maintaining per-project and global journal areas","Offline/local processing of embeddings and search"],"not_for":["Cloud-based shared journaling or collaboration across users","Storing sensitive data with strong security guarantees without additional hardening","High-throughput or multi-tenant deployments requiring robust API security"],"best_when":"Used by a single user locally (e.g., Claude Desktop) where privacy is achieved by keeping data on the host machine and where lightweight semantic search is sufficient.","avoid_when":"When you need strong authentication/authorization controls, centralized governance, or when the host environment cannot be trusted with sensitive personal data written to disk.","alternatives":["Other local RAG/journaling tools with explicit security controls","MCP servers that expose journaling/search via a more formal backend with access control","General-purpose note apps with local full-text/embedding search"],"af_score":60.2,"security_score":26.0,"reliability_score":22.5,"package_type":"mcp_server","discovery_source":["github"],"priority":"high","status":"evaluated","version_evaluated":null,"last_evaluated":"2026-03-30T13:33:28.393343+00:00","interface":{"has_rest_api":false,"has_graphql":false,"has_grpc":false,"has_mcp_server":true,"mcp_server_url":null,"has_sdk":false,"sdk_languages":[],"openapi_spec_url":null,"webhooks":false},"auth":{"methods":["No authentication described for MCP/stdio mode (tool runs locally via npx)."],"oauth":false,"scopes":false,"notes":"README describes running via `npx` with MCP stdio; no auth scheme, API keys, or access control is documented."},"pricing":{"model":null,"free_tier_exists":false,"free_tier_limits":null,"paid_tiers":[],"requires_credit_card":false,"estimated_workload_costs":null,"notes":"No hosted service or pricing details provided; appears to run locally."},"requirements":{"requires_signup":false,"requires_credit_card":false,"domain_verification":false,"data_residency":["Local machine only (implied by 'no external API calls required' / local processing)."],"compliance":[],"min_contract":null},"agent_readiness":{"af_score":60.2,"security_score":26.0,"reliability_score":22.5,"mcp_server_quality":78.0,"documentation_accuracy":72.0,"error_message_quality":0.0,"error_message_notes":null,"auth_complexity":95.0,"rate_limit_clarity":10.0,"tls_enforcement":20.0,"auth_strength":15.0,"scope_granularity":10.0,"dependency_hygiene":55.0,"secret_handling":40.0,"security_notes":"The design appears local/offline with no external API calls mentioned, which reduces data-in-transit risk. However, there is no documented authentication/authorization, and MCP runs locally via stdio; anyone who can invoke the MCP process can likely access/write the journal files. The README implies embeddings and journaling are stored on disk (including per-project and user home directories), so confidentiality depends on host security (filesystem permissions, device security). No guidance is provided on encryption-at-rest, file permissions, or safe handling of sensitive content.","uptime_documented":0.0,"version_stability":35.0,"breaking_changes_history":30.0,"error_recovery":25.0,"idempotency_support":"false","idempotency_notes":"No idempotency semantics described for write/journal operations; entries are timestamped and would likely create new files each call.","pagination_style":"limit-based (functions accept limit and days; no cursor described).","retry_guidance_documented":false,"known_agent_gotchas":["Running via `npx` means the agent environment must allow executing GitHub packages; network access may be needed for initial retrieval.","Local embeddings/index generation on startup can be slow; first-run latency may affect agent workflows.","Search results reference file paths; agents should treat returned paths carefully to avoid unintended file reads.","No documented rate limiting/auth—misconfiguration or misuse could lead to excessive disk writes or local resource exhaustion."]}}