dse120071750/m8m-harness-builder / skills/m8m/m8m-harness-builder/templates
__SKILL_NAME__
VerifiedM8M 3.1 product skill whose closed native source compiles into a displayable, executable milestone workflow.
SKILL_NAME
Invoke $__SKILL_NAME__ through agents/openai.yaml, the graph and public
contract authority. Each node owns agents/<milestone>.yaml and its complete
master prompt at references/<milestone>.md.
Author the full prompt first: role, goal, inputs and reference roles, instructions, constraints, and deliverable. Start execution and recovery with that whole prompt. Every milestone Markdown then lists numbered FlowSteps with descriptive actions and actual tools, followed by named outputs. Match agent bindings; leave no placeholder tool names.
Name new milestones milestone01, milestone02, etc.; keep IDs stable on edits.
Bind later inputs to earlier named outputs and explain their reference roles in
the prompt. Graph order controls execution.
Canonical flow.yaml is generated review output. Edit native source instead.
Generated tools remain BUILD_REQUIRED and non_runnable until implemented and
validated. planning/m8m-flowchart.md is an optional canvas projection.
Inputs are semantic context interpreted through the master prompt. Do not create or validate a milestone input schema. Preserve useful upstream bindings. Every milestone returns structured named outputs validated against its output schema.
Default to loop: none. Complete from actual named outputs and schema checks.
Do not add hashes, revisions, proof graphs, or automatic image reviews. A
checksum or review needs an explicit request or an external API requirement.
Keep one current chosen output per milestone.
Local coordination reuses existing tools without packaging. For an explicitly
packaged workflow, scripts/m8m_run.py uses the codebase launcher at
flowsteps/flows/<flow_id>/launch.py, which selects a digest-addressed runtime
release beside the harness. It must never import or invoke m8m-harness-builder.
This skill never pushes, publishes, deploys, or activates anything.
Builder 2.x run folders are untrusted import evidence.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 470k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |