Menu
Test-driven development with engram validation checkpoints. Write test first, watch it fail, write minimal code, store evidence at each phase.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Document architectural decisions as engraph ADRs. Store context, options considered, rationale, and consequences for queryable decision history.
Every commit must reference a valid engram task UUID. Covers format, validation, and why --no-verify is prohibited.
Orchestrator agent loop. Coordinates work across subagents using engram as shared state. Search before acting, dispatch by UUID, validate before finishing.
Plan feature implementation by creating an engram task hierarchy. Not a markdown plan — a queryable task tree any agent can retrieve.
Universal session end protocol. Run at the end of every agent session — closes open tasks, generates a handoff summary, syncs to remote, and validates state.
Universal session start protocol. Run at the beginning of every agent session — syncs from remote, opens a session, loads prior context, and surfaces the next action.
| name | engram-tdd |
| description | Test-driven development with engram validation checkpoints. Write test first, watch it fail, write minimal code, store evidence at each phase. |
Per-turn write rule: After every
bash(test run),edit, orwrite(code change), store evidence withengram_context_createorengram_reasoning_create+engram_relationship_createbefore ending the turn. Each phase below has an explicit write step — do not skip them.
NO PRODUCTION CODE WITHOUT A FAILING TEST FIRST
engram_ask query="<feature or function name> test"
engram_task_create title="TDD: <feature description>" priority="high"
# TASK_UUID
engram_task_update id=<TASK_UUID> status="in_progress"
Write one minimal test. Store evidence:
engram_context_create title="RED: <test_name>" content="Test: <test_name>\nLocation: <path>\nExpected: <behaviour>" source="<test-file>"
engram_relationship_create source_id=<TASK_UUID> source_type=task target_id=<CTX_UUID> target_type=context relationship_type=relates_to agent="<name>"
Run the test. Store output:
engram_reasoning_create title="RED verification: <test_name>" task_id=<TASK_UUID> content="Command: <cmd>\nOutput: <failure output>\nStatus: READY FOR GREEN"
engram_relationship_create source_id=<TASK_UUID> source_type=task target_id=<RSN_UUID> target_type=reasoning relationship_type=explains agent="<name>"
engram_validate
Write simplest code that passes. Store implementation:
engram_context_create title="GREEN: <test_name>" content="Implementation: <code>\nFiles: <list>" source="<file>"
engram_reasoning_create title="GREEN verification: <test_name>" task_id=<TASK_UUID> content="Tests: <N>/<N> pass\nStatus: READY FOR REFACTOR"
engram_validate
Only after green. No new behaviour. Store changes:
engram_reasoning_create title="REFACTOR: <test_name>" task_id=<TASK_UUID> content="Changes: ...\nTests still pass: YES\nNew behaviour: NO"
engram_validate
engram_task_update id=<TASK_UUID> status="done"
engram_next