Skip to content
Agentmart
Submit a tool

one-work-item-delivery-loop

Run the full loop from planned work or reported bugs to implementation, validation, and One Horizon write-back.

Skill by onehorizonai

IndexedMIT★ 2
Write a review

Chalk Labs

Need help deploying one-work-item-delivery-loop?

Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.

Book a call

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/onehorizonai-one-work-item-delivery-loop

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Run the full loop from planned work or reported bugs to implementation, validation, and One Horizon write-back. Use for prompts like "what do I have planned", "pick this up and implement it", "I found a problem", "fix all bugs assigned to me", "implement HubSpot lead sync", and "write this back to One Horizon". Requires One Horizon MCP.

Use cases: writing, sales, agent-building

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/one-work-item-delivery-loop

What's inside

Skill name
one-work-item-delivery-loop
Description
Run the full loop from planned work or reported bugs to implementation, validation, and One Horizon write-back. Use for prompts like "what do I have planned", "pick this up and implement it", "I found a problem", "fix all bugs assigned to me", "implement HubSpot lead sync", and "write this back to One Horizon". Requires One Horizon MCP.
License
MIT
Instructions
56 lines in SKILL.md · view on GitHub

Files (3)

  • SKILL.md
  • openai.yaml · agents/
  • logo.svg · assets/

SKILL.md (first lines)

# Work Item Delivery Loop

Use as few output tokens as possible while completing the task correctly. Write in plain English. Apply this to documents, progress messages, and final replies.

Deliver planned work or fixes through implementation, validation, and One Horizon write-back.

## Instructions

1. Identify the target with `list-planned-work`, `list-initiatives`, or `list-bugs`; fetch `get-task-details` before implementation. Confirm ambiguous initiative matches.
2. Use relevant tools and companion skills (`one-bug-triage-prep`, `one-initiative-summary`, recap/summarizer) before coding or write-back. Use text `Goals`, `Products`, `Skills` and metadata `goals`, `products`, `skills` to verify scope/linkage.
3. Implement, then validate with relevant checks. Never mark work complete without validation evidence.
4. After each delivered chunk, immediately use `update-bug`, `update-todo`, `update-initiative`, or completed `create-todo`, then comment. Do not stop at a status-only update after implementation. A run remains incomplete until write-back and requested initiative links are done.

In plan mode or when asked for a plan, explicitly include discovery, full details, implementation, validation, MCP status/comment write-back, and requested initiative links. Include per-bug implementation/write-back for batch fixes and matching/confirmation for initiatives. Do not skip required steps.

## Work types

Use initiatives for multi-day/roadmap planned work, bugs for unplanned defects, ongoing work for recurring owner-driven operations where supported, and Todos for small personal follow-ups. A completed initiative slice may be a completed Todo linked to its initiative.

## Specific flows

- **Fix all bugs assigned to me:** resolve user/team with `list-my-teams`; list active bugs filtered to the current assignee. For each: details → implement → validate → `update-bug` → comment with root cause, code changes, and current status. If not fixed, still update its current status and comment with the blocker. Stop if none match.
- **Implement an initiative:** list active initiatives; rank by title, then taxonomy/team; present top matches and confirm the selected target. Fetch details and verify label/scope fit. After implementation, immediately `create-todo` with `status: "Completed"`, `initiativeId`, workspace, and delivery description. Update initiative progress/status as appropriate and comment.
- **Connect to initiatives:** resolve each with `list-initiatives`, confirm ambiguity, then apply relation-capable tools. For Todos, set the primary `initiativeId` and use relations for additional links.

## Description and metadata edits

- Record progress in comments, never descriptions.
- Patch initiative descriptions with `patch-document`, `workspaceId`, initiative `taskId`, and precise `ops`: `replace_text`, `insert_before`, `insert_after`, `delete_text`. The server resolves or creates the linked document.
- Use `update-initiative` for metadata only: `title`, `status`, `assigneeIds`, `teamIds`, `taxonomyLabelIds`, `parentInitiativeId`. If both change, patch first, then update metadata.
- On stale/missing patch targets or anchors, fetch details and retry with corrected operations. Fetch again after mutation only if refreshed full content is needed.

## Write-back

Always use `source: "skill"` for `add-task-comment`.

Delivered work:

```markdown
**Changes**
- What changed: <delivered work>
- Why: <root cause or goal>
```

Research/planning/triage without delivery:

```markdown
## Update
- Summary: <finding or decision>
```

Report the target, action, MCP write-back, and linked initiative IDs briefly. If validation fails, keep work incomplete and write its current status/blocker. Missing task or label context requires details before proceeding.

Before declaring completion, ensure every completed bug/Todo/initiative has its status update (or completed Todo creation) and delivery comment in this run, and every requested initiative link is applied. Work must be implemented or explicitly blocked; perform missing write-back before finishing.

From onehorizonai/skills, MIT.

More from onehorizonai/skills

Signals

Stars
2
Forks
1
Open issues
0
Last commit (this folder)
Oct 9, 2026 (21 days ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from onehorizonai

Done for you

Rather not set it up yourself?

Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.

See how we build AI agents →Read the case studies

Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.

launches shipped
50+
launches shipped
media placements
500+
media placements
founders on every account
2
founders on every account

Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block

Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.

Get one-work-item-delivery-loop running in your business

Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.

Goes to the Chalk Labs founders, not the newsletter. Privacy

Need help deploying one-work-item-delivery-loop?

CallGet help
Cite this page

one-work-item-delivery-loop skill. Agent Mart. https://agentmart.dev/skills/onehorizonai-one-work-item-delivery-loop (last updated 2026-10-24)