wassnameandPi/OpenAI a4ed6cfbaa Show supervisor advice and restore reliable review lifecycle
Render exact instructions, restore monitoring/read-only tools on resume, report actual Pi idle state, reject stale approval views, and stop completed-plan timers. Ask for brief evidence-based judgment. Add real TUI rendering and lifecycle regressions; retain explicit limits on background state and unmeasured cost benefit.

Co-Authored-By: Pi/OpenAI <288921227+claudypoo@users.noreply.github.com>
2026-09-08 10:50:18 +08:00
2026-08-26 13:43:42 +08:00
2026-08-18 18:02:21 +08:00

pi-goals

Plan in one Pi session, then do the work there while a stronger visible Pi session supervises it.

How it works

  1. /goals <objective> creates .pi/plan/<session_id>-vN.md and enters read-only plan mode.
  2. Pi asks only material questions, writes the plan, and shows Ready / Refine / Edit / Cancel.
  3. Ready opens a second Herdr pane. The new Pi session explicitly forks the planning session and compacts that fork.
  4. The original session becomes the implementation worker. It keeps the full conversation and normal tools.
  5. The fork becomes a read-only supervisor. Worker views and supervisor instructions use a session-scoped mailbox under ignored .pi/goals-supervision/.
  6. Ready waits for the supervisor's durable readiness receipt; the worker does not begin before the fork has compacted and started.
  7. The supervisor compacts again when its context reaches 100k tokens.
  8. The supervisor records a private approval only after it sees a stopped worker, no active work, a clean commit, evidence, and saved verification output. CompleteGoal checks that approval against the exact plan block and Git tree before it ticks [x].

The two Pi sessions are visible. You can switch to the supervisor pane and talk to it directly. Supervisor instructions are shown in full, including in collapsed tool rows; ordinary messages and emitted thinking use Pi's display settings. The supervisor is prompted to give brief progress assessments and use judgment about when to intervene.

On resume, monitoring and read-only tools are restored. Periodic views report whether Pi is idle; they do not measure background jobs. Readiness is a startup receipt, not a continuous health check. Reviews stop after all goals are completed or cancelled, and both panes remain available. These mechanics are tested; useful judgment and savings from a cheaper worker still require a representative two-model run. -- Pi/OpenAI

Install

This branch requires Herdr 0.7.5 or newer and one Pi package:

pi install npm:@wassname2/pi-goals

For a local checkout:

pi -e .

Run Pi from the Git repository that the plan will change. Ready fails if the current directory is not inside a Git repository; this prevents approval from checking the wrong repository.

Commands

/goals <objective>          create a new plan
/goals model <model>        select the visible supervisor model
/goals model                use Pi's current default model
/goals clear                close the supervisor pane and disconnect the plan

/goals clear keeps the plan file. Starting another plan also keeps older versions.

Plan format

A goal is a checkbox line whose text starts with goal::

1. [ ] goal: Produce the report
   - subtle failure mode: the report exists but uses stale data
   - discriminator: the report cites the current input and the saved check confirms it
   - verify: `just verify`
   - evidence: (empty until sign-off)

The worker saves verification output in a nonempty repository file, adds that path to evidence, and commits it. The supervisor calls ApproveGoal with the inspected path; the worker then calls CompleteGoal with the exact goal text.

Development

npm test
npm run typecheck
npm run lint

test/rpc-review.test.ts runs the planning review flow through Pi's real RPC protocol with a local deterministic model. The Herdr launcher and visible supervisor bootstrap have focused tests; use a real Herdr session for the final two-pane check.

-- PI[gpt-5.6-sol]

S
Description
pi extension: Set goals in plan.md; a smart supervisor guides cheap worker subagents through long autonomous sessions until your goals are signed off, with every agents pane visible to you.
Readme
2.8 MiB
Languages
TypeScript 97.3%
JavaScript 2.7%