Long-running agent harness
Plan and scaffold work so an AI agent can make incremental, verifiable progress across many context windows. Flow: design doc (human + AI) → Initializer once (creates feature list, runbook, backlog, progress) → Coding agent in a loop until done. The implementation backlog defines one work unit per session and helps identify milestones for human check-in. Based on Anthropic: Effective harnesses for long-running agents.
Problem
Agents work in discrete sessions with no memory. Without structure they tend to: (1) try to do too much in one go and leave half-implemented, undocumented state; (2) declare the job done too early; (3) leave bugs or undocumented progress. A harness gives each new session a clear picture of state and one concrete unit of work.
Workflow overview
| Phase |
Who |
What |
| 0. Design doc |
Human + AI |
Iterate in chat (or doc) with the user to produce one design doc: scope, acceptance criteria, tech stack, run/verify steps. No implementation. |
| 1. Initializer (once) |
AI (sub-agent) |
Run the Initializer prompt. Reads _working/design.md. Creates/refreshes in _working/: feature list (YAML), runbook, implementation backlog, progress file, optional init script. Does not implement features or create project scaffolding (assumes existing repo). |
| 2. Coding loop |
AI (sub-agent each time) |
Run Coding agent sessions: read progress → pick one work unit from the backlog → implement → verify using runbook → update progress and feature list → commit. At milestones (e.g. end of a work-unit group): pause, check in with the human; when satisfied, re-run Initializer, then continue. Loop until all features pass or the user stops. |
The top-level agent invokes the Initializer and each Coding session as sub-agents, passing the relevant prompt each time (no need to copy prompts into .cursor/rules/). Use whatever mechanism Cursor provides to spawn a sub-agent or new task and pass the prompt content (e.g. the contents of the Initializer or Coding prompt file).
Phase 0: Design doc (human + AI)
- Goal: One design doc that defines scope, acceptance criteria, and enough context for the feature list and runbook.
- Process: Human and AI iterate (in chat or a shared doc). AI suggests structure and content; human refines. Optionally use the doc-coauthoring skill for the design doc.
- Output: Design doc at
_working/design.md. This is the input to the Initializer (which creates the feature list and other harness artifacts).
Two-phase execution (Initializer + Coding agent)
Initializer (run once)
Run the Initializer prompt (e.g. via sub-agent). It:
- Reads the design doc and any existing context.
- Creates or refreshes in
_working/: agent-runbook.md, implementation-backlog.md, agent-progress.md, feature_list.yaml, and optionally init_<app\>.sh.
- Does not implement features. Output: list of artifacts and the recommended first coding target.
Coding agent (every session, until done)
Run the Coding agent prompt (e.g. via sub-agent). Each session:
- Session start: If any of
_working/agent-progress.md, _working/agent-runbook.md, _working/implementation-backlog.md, _working/feature_list.yaml are missing → run Initializer first.
- Get bearings: Read
_working/agent-progress.md → _working/implementation-backlog.md → _working/feature_list.yaml → _working/design.md → _working/agent-runbook.md → recent git log.
- Pick one unit: At most one page/feature from the backlog that still has failing (or unimplemented) criteria.
- Implement: Implement only that unit. Add tests. Run lint/typecheck/tests from the runbook; fix failures.
- Feature list: Set
passes: true only for criteria verified this session.
- Handoff: Append a "Session N" block to progress; commit; leave repo runnable.
Loop: After each session, if not all features pass and the user has not stopped, run the Coding agent again (new sub-agent). Repeat until the next milestone or all features pass.
Milestone check-in: At important points (e.g. end of a logical work-unit group, or when the design doc defines a milestone), pause the coding loop. Check in with the human: summarise progress, show what’s done and what’s next. The human may revise the plan, adjust the current implementation, or change priorities. When the human is satisfied, re-run the Initializer (to refresh backlog and progress from any plan changes), then resume the Coding agent loop until the next milestone or completion.
Artifact layout
All harness artifacts live under _working/ (scratch space for the agent; see workspace rules).
| Location |
Purpose |
Who updates |
_working/design.md |
Scope, acceptance criteria, tech stack; input to feature list and Initializer. |
Human + AI (Phase 0). |
_working/agent-progress.md |
Current state, last session summary, next target, session history. |
Coding agent at end of each session. |
_working/agent-runbook.md |
How to start app, lint/typecheck, run tests; how to verify (e.g. unit, E2E, browser automation—exact commands and steps). |
Initializer; human when tooling changes. |
_working/implementation-backlog.md |
Ordered work units (one per session); route, acceptance criteria, dependencies. Defines milestones (e.g. end of a group). |
Initializer from design/context. |
_working/feature_list.yaml |
Acceptance criteria in YAML; each item has passes: true/false. |
Initializer creates and maintains; Coding agent sets passes: true only after verification per runbook. |
_working/init_<app>.sh (optional) |
One-command start (e.g. install deps, start servers). |
Initializer or human. |
Detailed templates and full prompt bodies live in references/REFERENCE.md.
Rules to enforce in prompts
- Incremental: One work unit per Coding session. If a unit is large, split across sessions and set "continue [unit X]" as next target.
- Clean state: Each session ends with a commit and a progress update. No partial broken scaffolding.
- Feature list discipline: Never remove or reorder items. Never set
passes: true without running the corresponding verification.
- Runbook as source of truth: Coding agent must use runbook commands for verify; do not guess or invent commands.
Feature list format (YAML example)
acceptance_tests:
- description: log in as admin redirects to app (root / dashboard)
category: auth
passes: false
steps:
- Start a new browser session
- Navigate to the login page
- Log in as the admin user
- Verify redirect to /
- description: invalid resource id shows 404 or clear error
category: edge-case
passes: false
steps: [...]
The feature list is YAML only; when updating passes, preserve the rest of the file structure.
Reference
Source: Michael-F-Bryan/mcc — distributed by TomeVault.
1---2name: long-running-agent-harness3description: Plans and structures large-scale work for AI agents across many sessions. Human and AI iterate to produce a design doc; run an Initializer sub-agent once to create feature list, runbook, and backlog in _working/; then repeatedly run a Coding sub-agent until all features pass. At milestones (e.g. end of a work-unit group), pause for human check-in, re-run Initializer, then continue. Prompts are passed to sub-agents when spawning (no copying into .cursor/rules). Use when planning multi-session agent work, long-horizon coding from a design, or handoff between coding sessions. Use when this capability is needed.4---56# Long-running agent harness78Plan and scaffold work so an AI agent can make **incremental, verifiable progress** across many context windows. Flow: **design doc (human + AI)** → **Initializer once** (creates feature list, runbook, backlog, progress) → **Coding agent in a loop** until done. The implementation backlog defines **one work unit per session** and helps identify **milestones** for human check-in. Based on [Anthropic: Effective harnesses for long-running agents](https://www.anthropic.com/engineering/effective-harnesses-for-long-running-agents).910## Problem1112Agents work in discrete sessions with no memory. Without structure they tend to: (1) try to do too much in one go and leave half-implemented, undocumented state; (2) declare the job done too early; (3) leave bugs or undocumented progress. A **harness** gives each new session a clear picture of state and one concrete unit of work.1314## Workflow overview1516| Phase | Who | What |17|-------|-----|------|18| **0. Design doc** | Human + AI | Iterate in chat (or doc) with the user to produce one design doc: scope, acceptance criteria, tech stack, run/verify steps. No implementation. |19| **1. Initializer (once)** | AI (sub-agent) | Run the Initializer prompt. Reads `_working/design.md`. Creates/refreshes in `_working/`: feature list (YAML), runbook, implementation backlog, progress file, optional init script. Does **not** implement features or create project scaffolding (assumes existing repo). |20| **2. Coding loop** | AI (sub-agent each time) | Run Coding agent sessions: read progress → pick **one work unit** from the backlog → implement → verify using runbook → update progress and feature list → commit. **At milestones** (e.g. end of a work-unit group): pause, check in with the human; when satisfied, re-run Initializer, then continue. **Loop until all features pass or the user stops.** |2122The top-level agent invokes the Initializer and each Coding session as **sub-agents**, passing the relevant prompt each time (no need to copy prompts into `.cursor/rules/`). Use whatever mechanism Cursor provides to spawn a sub-agent or new task and pass the prompt content (e.g. the contents of the Initializer or Coding prompt file).2324## Phase 0: Design doc (human + AI)2526- **Goal**: One design doc that defines scope, acceptance criteria, and enough context for the feature list and runbook.27- **Process**: Human and AI iterate (in chat or a shared doc). AI suggests structure and content; human refines. Optionally use the **doc-coauthoring** skill for the design doc.28- **Output**: Design doc at `_working/design.md`. This is the **input** to the Initializer (which creates the feature list and other harness artifacts).2930## Two-phase execution (Initializer + Coding agent)3132### Initializer (run once)3334Run the Initializer prompt (e.g. via sub-agent). It:35361. Reads the design doc and any existing context.372. Creates or refreshes in `_working/`: `agent-runbook.md`, `implementation-backlog.md`, `agent-progress.md`, `feature_list.yaml`, and optionally `init_<app\>.sh`.383. Does **not** implement features. Output: list of artifacts and the recommended first coding target.3940### Coding agent (every session, until done)4142Run the Coding agent prompt (e.g. via sub-agent). Each session:43441. **Session start**: If any of `_working/agent-progress.md`, `_working/agent-runbook.md`, `_working/implementation-backlog.md`, `_working/feature_list.yaml` are missing → run Initializer first.452. **Get bearings**: Read `_working/agent-progress.md` → `_working/implementation-backlog.md` → `_working/feature_list.yaml` → `_working/design.md` → `_working/agent-runbook.md` → recent git log.463. **Pick one unit**: At most one page/feature from the backlog that still has failing (or unimplemented) criteria.474. **Implement**: Implement only that unit. Add tests. Run lint/typecheck/tests from the runbook; fix failures.485. **Feature list**: Set `passes: true` **only** for criteria verified this session.496. **Handoff**: Append a "Session N" block to progress; commit; leave repo runnable.5051**Loop**: After each session, if not all features pass and the user has not stopped, run the Coding agent again (new sub-agent). Repeat until the next **milestone** or all features pass.5253**Milestone check-in**: At important points (e.g. end of a logical work-unit group, or when the design doc defines a milestone), **pause** the coding loop. Check in with the human: summarise progress, show what’s done and what’s next. The human may revise the plan, adjust the current implementation, or change priorities. When the human is satisfied, **re-run the Initializer** (to refresh backlog and progress from any plan changes), then **resume the Coding agent loop** until the next milestone or completion.5455## Artifact layout5657All harness artifacts live under **`_working/`** (scratch space for the agent; see workspace rules).5859| Location | Purpose | Who updates |60|----------|---------|-------------|61| **`_working/design.md`** | Scope, acceptance criteria, tech stack; input to feature list and Initializer. | Human + AI (Phase 0). |62| **`_working/agent-progress.md`** | Current state, last session summary, next target, session history. | Coding agent at end of each session. |63| **`_working/agent-runbook.md`** | How to start app, lint/typecheck, run tests; **how to verify** (e.g. unit, E2E, browser automation—exact commands and steps). | Initializer; human when tooling changes. |64| **`_working/implementation-backlog.md`** | Ordered work units (one per session); route, acceptance criteria, dependencies. Defines milestones (e.g. end of a group). | Initializer from design/context. |65| **`_working/feature_list.yaml`** | Acceptance criteria in **YAML**; each item has `passes: true/false`. | Initializer creates and maintains; Coding agent sets `passes: true` only after verification per runbook. |66| **`_working/init_<app>.sh`** (optional) | One-command start (e.g. install deps, start servers). | Initializer or human. |6768Detailed templates and full prompt bodies live in [references/REFERENCE.md](references/REFERENCE.md).6970## Rules to enforce in prompts7172- **Incremental**: One work unit per Coding session. If a unit is large, split across sessions and set "continue [unit X]" as next target.73- **Clean state**: Each session ends with a commit and a progress update. No partial broken scaffolding.74- **Feature list discipline**: Never remove or reorder items. Never set `passes: true` without running the corresponding verification.75- **Runbook as source of truth**: Coding agent must use runbook commands for verify; do not guess or invent commands.7677## Feature list format (YAML example)7879```yaml80acceptance_tests:81 - description: log in as admin redirects to app (root / dashboard)82 category: auth83 passes: false84 steps:85 - Start a new browser session86 - Navigate to the login page87 - Log in as the admin user88 - Verify redirect to /89 - description: invalid resource id shows 404 or clear error90 category: edge-case91 passes: false92 steps: [...]93```9495The feature list is **YAML only**; when updating `passes`, preserve the rest of the file structure.9697## Reference9899- Directory layout, artifact templates, and prompt bodies for sub-agents: [references/REFERENCE.md](references/REFERENCE.md).100- Anthropic: [Effective harnesses for long-running agents](https://www.anthropic.com/engineering/effective-harnesses-for-long-running-agents).101102---103> Source: [Michael-F-Bryan/mcc](https://github.com/Michael-F-Bryan/mcc) — distributed by [TomeVault](https://tomevault.io).104<!-- tomevault:4.0:skill_md:2026-06-16 -->
tomevault-io