Harnesses and roles
Every agent harness Gantry drives, the environment keys that configure each one, and which role runs which stage of the pipeline.
The harnesses
Agent harnesses — 4 rows exported from
src/engine/harness/mod.rs
Claude
picked with c
| field | environment key | if that is unset |
|---|---|---|
| bin | GANTRY_CLAUDE_BIN | GANTRY_CLAUDE |
| model | GANTRY_CLAUDE_MODEL | GANTRY_MODEL |
| small_model | GANTRY_CLAUDE_SMALL_MODEL | GANTRY_COMMIT_MODEL |
| args | GANTRY_CLAUDE_ARGS | the harness's own default |
OpenCode
picked with o
| field | environment key | if that is unset |
|---|---|---|
| bin | GANTRY_OPENCODE_BIN | the harness's own default |
| model | GANTRY_OPENCODE_MODEL | the harness's own default |
| small_model | GANTRY_OPENCODE_SMALL_MODEL | the harness's own default |
| args | GANTRY_OPENCODE_ARGS | the harness's own default |
Codex
picked with x
| field | environment key | if that is unset |
|---|---|---|
| bin | GANTRY_CODEX_BIN | the harness's own default |
| model | GANTRY_CODEX_MODEL | the harness's own default |
| small_model | GANTRY_CODEX_SMALL_MODEL | the harness's own default |
| args | GANTRY_CODEX_ARGS | the harness's own default |
Gemini
picked with g
| field | environment key | if that is unset |
|---|---|---|
| bin | GANTRY_GEMINI_BIN | the harness's own default |
| model | GANTRY_GEMINI_MODEL | the harness's own default |
| small_model | GANTRY_GEMINI_SMALL_MODEL | the harness's own default |
| args | GANTRY_GEMINI_ARGS | the harness's own default |
The roles
Harness roles — 4 rows exported from
src/domain/mod.rs
| role | stages it runs |
|---|---|
| plan | Brief, Plan, Test setup, Replan |
| build | Execute, Fix, Investigate, Troubleshoot, Resolve |
| review | Review |
| util | Test, Merge, Support, Done |
The stages
Pipeline stages — 14 rows exported from
src/domain/mod.rs
| stage | shown as | description | run by |
|---|---|---|---|
| environment-build | Brief | Writing the task's environment brief, between its plan and its gate-builder | plan |
| plan | Plan | Planning the build | plan |
| gate-build | Test setup | Setting up the test command | plan |
| execute | Execute | Implementing | build |
| gate | Test | Testing | util |
| fix | Fix | One bounded repair after a red gate. Bundled `map` still schedules it; other recovery paths use `troubleshoot` | build |
| investigate | Investigate | Diagnosing a failed task and writing a report without repairing it. No bundled pattern schedules it; `troubleshoot` replaced it, and it is kept so older journals still render | build |
| troubleshoot | Troubleshoot | Troubleshooting the failed task | build |
| review | Review | Reviewing | review |
| replan | Replan | Re-planning the remaining work | plan |
| merge | Merge | Merging the run branch back to the run's pinned home branch | util |
| resolve | Resolve | Resolving the merge conflict so both sides survive, and where they cannot, the home branch wins | build |
| support | Support | Run-level work outside the task pipeline: a blocker raised for an operator answer, and the run-record bookkeeping a pattern declares here | util |
| done | Done | The run finished green and merged — the terminal card | util |