MCP Server
SpecShip runs as a Model Context Protocol server. Start it with:
specship serve --mcpAgents configured by the installer launch this automatically. When a .specship/ index exists, the agent uses the tools below.
Core (always available): specship_callees · specship_callers · specship_drifted · specship_explore · specship_files · specship_fitness · specship_impact · specship_link_assert · specship_link_verify · specship_maintainability · specship_node · specship_search · specship_spec · specship_status
JIRA integration (opt-in, specship install --with-jira): specship_jira_issue · specship_jira_issues · specship_jira_pick · specship_jira_publish · specship_jira_start · specship_jira_track · specship_jira_transition
Designer integration (opt-in + experimental, --with-designer): designer_ask · designer_handoff · designer_list · designer_prompt · designer_session · designer_snapshot
Code-graph tools
Section titled “Code-graph tools”| Tool | Purpose |
|---|---|
specship_explore | Primary. Return the verbatim source of the symbols relevant to a question — a plain question or a bag of symbol names both work — grouped by file, plus a relationship map and the call path among named symbols (following dynamic-dispatch hops like callbacks, React re-render, and interface→impl that grep can’t). Also surfaces matching human-confirmed domain facts and a “Prior work” section (past sessions that edited these files) inline. Usually the only call an agent needs. |
specship_search | Find symbols by name across the codebase (just locations). |
specship_node | Get one symbol’s full body + its caller/callee trail; returns every overload’s body in one call for an ambiguous name. |
specship_callers | Find what calls a function. |
specship_callees | Find what a function calls. |
specship_impact | Analyze what code is affected by changing a symbol. |
specship_files | Get the indexed file structure (faster than a filesystem scan). |
specship_status | Check index health and statistics. |
Spec tools
Section titled “Spec tools”| Tool | Purpose |
|---|---|
specship_spec | Fetch a spec by ID with its currently linked code. |
specship_link_assert | Declare an implements / tests / documents link after editing. (Mutating — prompts for permission.) |
specship_link_verify | Mark a link’s verification pass/fail. (Mutating — prompts for permission.) |
specship_drifted | Pull the drift queue (drifted / broken / orphaned links). |
JIRA tools
Section titled “JIRA tools”The issue-driving flow. Connect first with specship jira configure. See the JIRA integration guide.
| Tool | Purpose |
|---|---|
specship_jira_issues | List the issues assigned to you (resolved from your token) — key, summary, status, type. Optional project filter. |
specship_jira_issue | Fetch one issue by key — summary, description, status, type. |
specship_jira_pick | Draft a well-formed SpecShip spec from an issue under specs/. Idempotent — re-picking updates the spec in place. |
specship_jira_start | Run the spec-implement workflow on the generated spec; pauses at its plan/approve gate. |
specship_jira_track | Read-only table joining each picked issue’s SpecShip work-state with its live JIRA status. |
Design tools
Section titled “Design tools”Merged in from the standalone designer MCP, these drive claude.ai/design over a debug Chrome. macOS only; they only launch Chrome when invoked. See Design-to-code.
| Tool | Purpose |
|---|---|
designer_session | Open or attach a Claude Design session. |
designer_prompt | Send a design prompt and iterate on variants. |
designer_ask | Ask a question about the current design. |
designer_list | List the files/variants in the session. |
designer_snapshot | Snapshot a design’s source byte-for-byte to disk. |
designer_handoff | Produce a handoff bundle for the implement workflow. |
How agents should use it
Section titled “How agents should use it”SpecShip is the pre-built search index. For “how does X work?”, architecture, trace, or where-is-X questions, an agent should answer in a handful of SpecShip calls and stop — typically with zero file reads — rather than re-deriving the answer with grep + Read. A direct SpecShip answer is a handful of calls; a grep/read exploration is dozens.
The installer writes this guidance into each agent’s instructions file automatically.