Pin a session goal and keep pi anchored to it. The agent is re-prompted at every settled idle boundary until the objective is genuinely achieved — with safety limits, usage accounting, and evidence-gated completion.
Part of the Pify suite. Install with pify install goal or pi install npm:@pify/goal.
/goal ship structured logging across the auth module # set + start
/goal # status (objective, state, usage)
/goal pause # stop continuation
/goal resume # continue (also resets safety limits)
/goal <new objective> # supersede the objective mid-flight
/goal clear # drop the goal
/goal audit on # verify every completion claim (v0.3)
/goal steps # progress through an ordered goal (v0.4)
A Codex-style footer indicator tracks the state: 🎯 Pursuing goal (5m 12s) → 🎯 Goal achieved (23m).
- One goal per session, persisted as session entries — it survives
/reload, forks, branch switches, and compaction. Reopening a session with an active goal never auto-runs; the goal resumes after your next message. - Continuation at the settled boundary: after queued work, retries, and compaction finish (
agent_settled), the extension queues one hidden follow-up prompt. Nothing is injected into the system prompt, so the provider request prefix stays stable and prompt caching keeps working. - The user owns intent: the agent has no tool to create or resume goals. It reports through three evidence-gated tools —
goal_complete(requires a completion audit + evidence),goal_blocked(requires the same blocker recurring 3+ turns),goal_wait(external events). Your next message automatically reactivates a blocked/waiting goal. - Sisyphus mode (v0.4): write the objective as a list — one step per line, or separated by
;— and the goal is worked in order. Every continuation prompt marks which step is current (the list itself stays visible — it is your objective),goal_step_done(evidence)advances the cursor one step forward and only forward, andgoal_completeis refused while steps remain: an ordered goal is finished when its list is, not when the agent feels done with the interesting part. Prose objectives are unaffected — a sentence with commas is not a plan, so it is never parsed as one. - Independent completion audit (v0.3, opt-in via
/goal audit on): the agent that did the work is the worst judge of whether it is done — it already believes it is. With the audit on, everygoal_completeclaim is handed to a second agent with read-only tools, no extensions, and instructions to disprove the claim against the repository as it stands. A rejection keeps the goal active and hands the agent a specific reason to fix. An auditor that times out or answers unreadably is recorded as inconclusive and lets the completion through — a broken auditor must never be able to trap the agent in a loop it cannot exit. - Errors that repeating cannot fix stop the goal (v0.5): a goal drives its own turns, so an expired key, an empty balance, a context overflow, or a model the account cannot reach would otherwise be retried on every continuation — forever, at the cost of a request each time. Those pause the goal with the reason; rate limits, overloads, and dropped connections deliberately do not, because riding those out is what the loop is for. (Classification from
pi-code.) - Safety limits: after 20 automatic turns without user input, or 3 consecutive tool-free turns with identical output (SHA-256 over normalized visible text), the goal pauses for review instead of looping.
- Interrupts respected: pressing Esc during a goal turn pauses the goal — the agent never barrels on after you grabbed the wheel.
- Prompt injection hygiene: the objective travels wrapped in
<untrusted_objective>tags, XML-escaped, marked as data — it cannot escalate into instructions. - Usage accounting: tokens and wall time per goal, shown in the footer,
/goalstatus, and every continuation prompt.
MIT © Pify maintainers