{"id":"tma1-ai-devtap","name":"devtap","homepage":null,"repo_url":"https://github.com/tma1-ai/devtap","category":"automation","subcategories":[],"tags":["ai-coding","mcp","devtools","cli","build-logs","debugging","go","greptimedb"],"what_it_does":"devtap captures stdout/stderr from local build/dev commands, stores pending output per AI adapter, and exposes those captured build errors to AI coding tools via an MCP-compatible stdio server (or via a CLI skill fallback). It can also fan out output to multiple agents and supports cross-machine workflows using an optional GreptimeDB-backed session store.","use_cases":["Automate copy/paste of build/test errors into AI coding sessions","Support parallel debugging with multiple AI agents consuming the same captured failures","Feed CI/remote build logs into a local AI coding tool session","Speed up iterative fix loops for local dev servers and watch tasks"],"not_for":["Replacing real observability/structured log pipelines","Handling untrusted remote log ingestion as a security boundary","Acting as a general agent-to-agent messaging system without local constraints"],"best_when":"You run AI coding tools that support MCP (or compatible adapters) and you want a tight feedback loop from build/dev stderr/stdout into automated error-fixing sessions, including multi-agent and cross-machine scenarios.","avoid_when":"You need hard guarantees about data sanitization against malicious log content, or you require a networked hosted API/web service rather than a local MCP/stdio workflow.","alternatives":["Manually paste logs into the coding tool","Use a dedicated log shipper + indexing (e.g., ELK/Opensearch) and prompt from aggregated error excerpts","Gremlin-style/agent workflows that poll build logs directly (tool-specific integrations)","Custom wrappers around your build system to extract errors and feed them to the agent"],"af_score":65.8,"security_score":54.5,"reliability_score":32.5,"package_type":"mcp_server","discovery_source":["github"],"priority":"high","status":"evaluated","version_evaluated":null,"last_evaluated":"2026-03-30T15:39:37.197309+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 for local stdio MCP (implicit trust in local user/process).","GreptimeDB credentials via environment variables (DEVTAP_GREPTIMEDB_USERNAME/DEVTAP_GREPTIMEDB_PASSWORD) when using the GreptimeDB store."],"oauth":false,"scopes":false,"notes":"devtap is primarily local-process orchestration; authentication is only relevant for the optional GreptimeDB backend."},"pricing":{"model":null,"free_tier_exists":false,"free_tier_limits":null,"paid_tiers":[],"requires_credit_card":false,"estimated_workload_costs":null,"notes":"Open source (MIT) CLI tool; costs are from the AI coding tool/model you use and optionally your own GreptimeDB infrastructure."},"requirements":{"requires_signup":false,"requires_credit_card":false,"domain_verification":false,"data_residency":["Local file store under ~/.devtap (file backend)","Self-hosted GreptimeDB instance when using the GreptimeDB backend"],"compliance":[],"min_contract":null},"agent_readiness":{"af_score":65.8,"security_score":54.5,"reliability_score":32.5,"mcp_server_quality":70.0,"documentation_accuracy":75.0,"error_message_quality":null,"error_message_notes":"Mentions troubleshooting behaviors (e.g., status/drain, session pick) and drain warnings for unreachable sources; no explicit MCP error schema documented.","auth_complexity":90.0,"rate_limit_clarity":25.0,"tls_enforcement":80.0,"auth_strength":55.0,"scope_granularity":20.0,"dependency_hygiene":45.0,"secret_handling":70.0,"security_notes":"Strengths: designed for local stdio MCP (no network sockets) and claims no telemetry; GreptimeDB credentials are provided via environment variables for the optional store. Concerns: there is no evidence of fine-grained auth scopes (local trust model), and rate limiting/abuse controls are not discussed. Dependency hygiene and CVE status cannot be confirmed from the provided README/metadata alone.","uptime_documented":0.0,"version_stability":40.0,"breaking_changes_history":30.0,"error_recovery":60.0,"idempotency_support":"false","idempotency_notes":"Drain/consumption model suggests queued consumption; idempotency semantics are not clearly documented as safe retries for the same drain call.","pagination_style":"none","retry_guidance_documented":false,"known_agent_gotchas":["MCP tool calls require running `devtap install --adapter <adapter>` in the project root and restarting the AI tool session.","Cross-machine/multi-source drain can emit warnings when some sources are unreachable; agents should handle partial availability.","Filtering options (regex/sql) are convenience features and not a security boundary; malicious content handling is not guaranteed."]}}