MCP Server¶

DSAgt exposes its capabilities through a single MCP server, dsagt-server, configured in the per-agent runtime file (.mcp.json for Claude Code, goose.yaml for Goose, etc.) and launched automatically when the agent starts. It combines four capabilities — a code registry, a knowledge base, explicit memory, and skill discovery — behind one process with one shared embedder and one ChromaDB.

The 20 tools split across four concerns, all on the one process.

Registry tools (8)¶

Code registration, execution helpers, dependency installation, and pipeline reconstruction. See Provenance for how registered codes are captured.

Tool	Description
`search_registry`	Semantic search over registered + built-in code specs
`get_registry`	List every registered code with its MCP-compatible schema
`save_code_spec`	Register a CLI code as `codes/<name>/SKILL.md` (executable auto-wrapped with `dsagt-run` + `uv run --with`), mirrored into the agent's native skills dir
`install_dependencies`	Install a code's Python dependencies via `uv run --with`
`run_command`	Execute a command to capture its help/usage output
`read_file`	Read a text file from disk
`http_request`	Fetch documentation or an API spec over HTTP(S)
`reconstruct_pipeline`	Render `trace_archive/` as a dependency-ordered bash script or Snakemake workflow

Codes are markdown files with YAML frontmatter under <project>/codes/. Executables are wrapped with dsagt-run for provenance and uv run --with for Python dependencies.

Knowledge tools (5)¶

Semantic search and ingestion over indexed document collections. See the Knowledge Base for the retrieval model.

Tool	Description
`kb_search`	Hybrid semantic search across one or more collections (optional metadata, regex, and substring filters)
`kb_ingest`	Index a folder as a new collection — returns a `job_id`; poll `kb_job_status` until complete
`kb_append`	Add documents to an existing collection (background job)
`kb_list_collections`	List collections with their embedding model and vector DB
`kb_job_status`	Check the status of a background ingest/append job

Memory tools (2)¶

User-confirmed facts that persist across sessions. See Memory.

Tool	Description
`kb_remember`	Store a user-confirmed fact as an explicit memory (`supersedes` to replace an outdated one)
`kb_get_memories`	Load active explicit memories for this project (call at session start)

Skill tools (5)¶

Discover, install, and author agent skills, and manage external skill sources. See Skills.

Tool	Description
`search_skills`	Search installed skills + the external corpus (corpus hits are tagged `[catalog]` in results)
`install_skill`	Copy a skill from the corpus into `<project>/skills/` and mirror it into the agent's native skills dir
`save_skill`	Register an agent-authored skill into `<project>/skills/<name>/SKILL.md`
`add_skill_source`	Enable + index an external skill source (a known name or a Git URL) into the searchable corpus
`list_skill_sources`	List known and synced external skill sources