## Thinking Path > - Paperclip manages work for AI agents. > - Planning guidance helps agents choose owners and dependencies. > - The runtime skill favors few tasks, but the catalog skill requires a child-task breakdown. > - Both add repeated process instructions that can distract from the requested outcome. > - This change keeps the ownership and dependency rules and removes the required matrix and repeated checklist. > - A bounded Product E2E comparison measures saved outcomes and task handoffs before qualification. ## Linked Issues or Issue Description Refs #11057. Related measurement work: #15218. **What existing behavior does this improve?** Planning and delegation through the runtime plan-to-tasks and bundled task-planning skills. **Current behavior** The two skills contain about 1,900 words and conflicting guidance on whether plans require child tasks. **Proposed behavior** Keep cohesive work with one owner. Split only for a real owner, parallel output, dependency, independent review, or follow-up lifecycle. Preserve existing authorization and planning mechanics. ## What Changed - Shorten both skills to about 400 words combined. Preserve their keys and installed-version behavior. - Remove the duplicate operational-skill pointer and regenerate affected source metadata. - Add twelve explicit Product E2E cells: four scenarios with current, short and disabled planning skills. - Use the current task composer and actual create-response ID; calibrate public skill APIs and browser creation without providers. - Eliminate an observed collision in chat-test company prefixes with a per-suite sequence. - Grade saved documents, exact author/run attribution, child count, prerequisite execution order, review boundaries and completion handoffs. - Retain current skill bytes and report source, selections, run accounting and failures. ## Verification - `pnpm test:e2e:runner:typecheck`: pass. - `pnpm test:e2e:runner:unit`: 1,287 Vitest tests and 128 Node checks pass. - `pnpm test:e2e:runner -- --list --suite plan-task-guidance`: twelve local Codex cells. - Archived current skills match master `72ff3a9f27e581a27acb49771e8658bbb0bbaa47` exactly. - Corrected fixture: three real public-API/database calibrations pass with zero provider runs; all 35 evaluator checks and Product E2E typecheck pass. - Setup campaign [37399550253](https://github.com/paperclipai/paperclip/actions/runs/37399550253) was canceled after source review found unsupported bundled edits and automatic core reinstallation. Its paid-cell step was skipped: zero provider runs, no behavioral grade. - The next setup [37401094799](https://github.com/paperclipai/paperclip/actions/runs/37401094799) failed before task creation on the old title-field selector: zero actual runs, original FAIL retained, cleanup passed. A real browser/API calibration of the new helper passes with paused non-provider agents and zero runs. - Full local typecheck/build pass. Full local tests retain one unchanged five-minute Git streaming timeout (also fails isolated), 9,591 passes and 5,796 skips. CI's chat failure was a proven random fixture-prefix collision; five affected cases pass after the test-only repair. - Paid behavior comparison and new-head CI/review remain pending. This PR remains a draft. ## Risks - The shorter text may change delegation decisions. Live outcomes are not yet qualified. - The initial comparison uses one profile and one attempt per cell. It cannot establish cross-model reliability or cost trends. - Disabled means unassigned company-owned copies; the company library remains discoverable. This does not qualify global removal, automatic accepted-plan wiring changes, or installed-copy migration. - Skill availability does not prove a model read or cognitively used it. - No provider/tool protocol, permission, timeout or runtime lifecycle behavior changes in production. ## Model Used OpenAI Codex (GPT-6), with repository inspection, code editing and tool use. The exact backend model ID and context-window size are not exposed in this session. The declared eval model is native Codex `gpt-5.6-sol`. ## Checklist - [x] I have included a thinking path that traces from project context to this change - [x] I have specified the model used (with version and capability details) - [x] I have checked ROADMAP.md and confirmed this PR does not duplicate planned core work - [x] I have searched GitHub for duplicate or related PRs and linked them above - [x] I have either (a) linked existing issues with `Fixes: #` / `Closes #` / `Refs #` OR (b) described the issue in-PR following the relevant issue template - [x] I have not referenced internal/instance-local Paperclip issues or links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip` URLs) - [x] My branch name describes the change (e.g. `docs/...`, `fix/...`) and contains no internal Paperclip ticket id or instance-derived details - [ ] I have run tests locally and they pass - [x] I have added or updated tests where applicable - [x] I have updated relevant documentation to reflect my changes - [x] I have considered and documented any risks above - [ ] All Paperclip CI gates are green - [ ] Greptile is 5/5 with no open P2s, recommendations, or follow-ups - [x] I will address all Greptile and reviewer comments before requesting merge --------- Co-authored-by: Paperclip <noreply@paperclip.ing>
4.8 KiB
name, description, key, recommendedForRoles, tags
| name | description | key | recommendedForRoles | tags | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| task-planning | Turn a Paperclip issue or request into a structured implementation plan with child task graph, blockers, owners, and acceptance criteria, then save it as the issue `plan` document. | paperclipai/bundled/paperclip-operations/task-planning |
|
|
Task Planning
Produce implementation plans that the Paperclip executor can actually run: explicit child issues, real blockers, named owners, and a defined acceptance bar. Avoid plans that read well but cannot be split into work.
When to use
- An issue asks you to "plan", "scope", "break down", "design the rollout", "propose the work", or similar.
- A user wants a written plan before approving implementation.
- A manager needs to delegate non-trivial work and the shape of the work is not obvious yet.
- You inherited an issue too large to deliver in one heartbeat and need to split it.
When not to use
- The issue is a single small change you can ship in the same heartbeat. Just ship it.
- The issue is forensic ("why did this break"). Use a diagnosis skill first; plan only after the root cause is named.
- A current
plandocument already exists and the change is minor. Update that document; do not start fresh.
Outputs
- An updated issue document with key
plan(markdown). - A short comment on the issue that links to the plan document and names the next action.
- Where the plan requires approval, an issue-thread interaction of kind
request_confirmationbound to the latest plan revision.
Do not create implementation subtasks until the plan is accepted.
Plan structure
Required sections, in order:
- Goal — one paragraph. What changes for the user, the operator, or the system once this work lands.
- Context reviewed — bullet list of documents, files, and prior issues you read. Lets reviewers spot missing inputs.
- Constraints and non-goals — what must hold (compatibility, security, performance) and what this plan deliberately will not do.
- Approach — the chosen path, with a short rationale. If you considered alternatives, name them and why you rejected them.
- Work breakdown — ordered list of child issues. Each child has:
- Title in imperative form.
- Owner specialty (Engineer, QA, Designer, Security, DevRel, Manager, etc.).
- Scope and deliverables.
- Acceptance criteria.
- Blocks/blocked-by relationships expressed by phase letter or child title.
- Acceptance — the bar for the parent issue. How the user knows the whole thing is done.
- Risks and mitigations — short list. Skip if there are none.
- Deferrals — what is intentionally pushed to follow-up issues, with why.
Rules of thumb for splitting
- One child issue, one specialty. If two specialties have to coordinate inside the same issue, split it.
- One child issue, one acceptance verdict. If a reviewer would say "this is half done", split it.
- A child must be checkout-able by the owner from its title and description alone. Reviewers should not have to re-read the parent plan to understand a child.
- Order children by real blocker chains, not by author preference. Parallel children should explicitly say
blockers: none. - Avoid
polishorcleanupchild issues without acceptance criteria — they never close.
Filing the plan
Use the Paperclip API to write the plan document, then comment:
PUT /api/issues/{issueId}/documents/planwith the markdown body. Ifplanalready exists, include the latestbaseRevisionId.POST /api/issues/{issueId}/commentswith a short summary that links the plan:/<prefix>/issues/<issue-id>#document-plan.- If approval is required:
POST /api/issues/{issueId}/interactionswithkind: request_confirmation,targetRevisionIdset to the new plan revision,continuationPolicy: wake_assignee, andidempotencyKey: "confirmation:{issueId}:plan:{revisionId}". - Set the issue to
in_reviewafter creating the confirmation. Stay assigned so the acceptance wakes the planner.
When the plan is accepted, see the companion skill for converting accepted plans into Paperclip executable tasks. Key requirements covered there: produce a compact task matrix (task, owner, initial status, blockers); encode every hard dependency as blockedByIssueIds — parent/child nesting alone does not block execution; and verify the created issue graph before closing the source planning issue.
Anti-patterns
- Plan disguised as a description edit. Use the
plandocument. - "Phases A–Z" with no work breakdown inside the phases.
- Children with descriptions that say "see parent" — they fail at delegation time.
- Acceptance written as "code review approval". Reviewers need a behavior bar, not a process bar.
- Plans that bury blocker chains in prose. Use explicit blocked-by lines.