increased autonomy in obot.agent

Goal issue #73 · slug autonomy

14/50 member issues closed

🟢 Ready for auto 1

obot can implement these unattended (label auto).

#132 roadmap hierarchy — every issue reachable from a goal (Backlog · whether the plan matches reality)

🟠 Needs steering 13

awaiting @jwildfire input or review (label draft).

#58 ideas-triage v2 — stronger model, richer context, bias-to-promote (Backlog · infrastructure · ai · whether the plan matches reality)
#87 weekly goal review — freshness sweep proposing linkages, labels, and orphan candidates (Requirement Gathering · infrastructure · ai · whether the plan matches reality)
#91 fast session startup — sub-10-second responsive /session-init (Design · infrastructure · how a session behaves)
#117 obot's Chrome tab sprawl — name, reuse, and clean up the tabs obot opens (Backlog · infrastructure · how a session behaves)
#123 release scaffolding — news record, draft notes, semver milestones, RC PR framework, and the merge gate (Backlog · infrastructure · how work reaches you)
#127 /oneoff — a single-agent lane that skips session startup but still logs to the roadmap (Requirement Gathering · infrastructure · how a session behaves)
#128 assignee and reviewer conventions for autonomous work — obot holds what's in flight, @jwildfire reviews (Backlog · infrastructure · how work reaches you)
#146 durable todo list — one permanent list per goal and overall, updated not rewritten (Requirement Gathering · infrastructure · ai · the surfaces you read)
#147 concurrent sessions — autonomous runs alongside @jwildfire's and each other (Requirement Gathering · infrastructure · ai · how a session behaves)
#148 fast, responsive /session-wrapup — the closing bookend (Requirement Gathering · infrastructure · ai · how a session behaves)
#122 scheduled autonomous sessions — the A2 promotion (recurring goal-work loop) (Requirement Gathering · infrastructure · how a session behaves)
#170 session hub — session reports as diary tabs (Backlog · the surfaces you read)
#178 audit stage history — re-gate the missing-design check to the moment work starts (Backlog · ai · whether the plan matches reality)

Other open members 22

no readiness label yet.

#48 idea queue — Siri/Reminders + hub Ideas Discussions intake with obot triage (Backlog · infrastructure · ai · whether the plan matches reality)
#69 Reviewer Notes section in PR template (Backlog · infrastructure · how work reaches you)
#76 obot GitHub App avatar — Orange's photo (Requirement Gathering · infrastructure · how work reaches you)
#157 Navigator — standing roadmap-state verifier (file-writing only) (Backlog · the officer and the workers)
#152 Four documents disagree on whether an increment PR should be a GitHub draft (infrastructure · ai · how work reaches you)
#94 Show release summaries on the roadmap page — recent releases render as an em-dash (infrastructure · the surfaces you read)
#180 the Operations Dashboard — my todo list, with blockers, where I answer decisions (Unstaged · infrastructure · ai · the surfaces you read)
#181 Navigator phase 2 — working-set verification (split from #157) (Unstaged · the officer and the workers)
#185 the first mate — in-flight supervision of worker agents (Backlog · the officer and the workers)
#186 standing sessions survive compaction — continuous state, an explicit setting, and a silent-loss check (Backlog · infrastructure · ai · how a session behaves)
#194 every worker agent gets a permanent W0001 identity, allocated once and never recycled (Review · the officer and the workers)
#195 the Navigator as operating officer — asks become requirements, and delivery is judged at closeout (Backlog · the officer and the workers)
#184 the worker closeout check — every agent finishes into a PR, a question, or a config request (Backlog · the officer and the workers)
#199 the agent roster — every agent with its id, status, cost and roadmap impact (Development · infrastructure · ai · the officer and the workers)
#200 roadmap discipline is measured — the sweep checks all seven repos, and the audit stops reporting stale (Development · whether the plan matches reality)
#205 a real "since you last looked" — record page views locally, or say plainly that we cannot (Development · infrastructure · the surfaces you read)
#206 the priced usage artifact refreshes on a cadence, or every surface says how old it is (Development · infrastructure · the surfaces you read)
#212 a worker that stops — finished, stalled or dead — wakes the Navigator (Backlog · infrastructure · the officer and the workers)
#215 a requirement says whose decision it carries — an agent's inference must not read as his approval (Backlog · how work reaches you)
#218 the dashboard reads like a news feed, not an audit log — rebuild the sessions and Navigator pages (Development · infrastructure · the surfaces you read)
#220 his queue holds exactly three buckets — RC, decision, config — and the remainder is ours (Requirement Gathering · the surfaces you read)
#223 every surface renders honestly on a machine with no history — before the move, not after (Backlog · infrastructure · the surfaces you read)

✅ Done 14

closed members.

#18 autonomous obot operations — scheduled and event-driven agent runs (Released · infrastructure · ai)
#61 reaction feedback and trigger reliability for the Ideas triage pipeline (Released · infrastructure)
#70 Mark PRs ready for review and request @jwildfire review when ready
#92 nightly roadmap audit — rules, confidence-scored findings, and a one-click apply loop (Released · infrastructure · ai)
#3 obot GitHub App for agent and automation access (Released · infrastructure · ai)
#24 session hub — live session dashboard + wrapup session report (Released · infrastructure · ai)
#31 roadmap transparency — requirements tied to session goals, releases, and live status (Released · infrastructure)
#77 chat with the orchestrator from the live session dashboard (Released · infrastructure · ai)
#109 trigger audit fixes from the page — one click, live run status, decisions logged in place (Released · infrastructure · ai)
#140 one write-policy decision per repo — unify merge tiers and autonomy grants into profiles (Released · infrastructure)
#198 no inline bold in artifacts — emphasis is structural, or it is a callout (Released)
#202 the roadmap page redesign — a design spike he can react to (Released · infrastructure)
#203 one information architecture, two audiences — symmetry between the private dashboard and the public hub (Backlog · infrastructure · the surfaces you read)
#211 rebuild the roadmap page on D0018 — the queue in front, the catalog behind (Development · infrastructure)

Direction

Standing goal — this issue stays open; closing it = retiring the goal. Direction and membership are maintained by @jwildfire; autonomous sessions propose changes as comments and never edit this issue or its sub-issue links.

Goal

Expand how much of the roadmap obot.agent can carry unattended, while keeping @jwildfire's review gates exactly where the operating contract puts them. Requested in discussion #54; this continues the direction set by #18 (autonomy design, --auto v1), whose maiden A1 run completed 2026-07-24.

Scope and boundaries

In scope, roughly in order: hardening the --auto lane from the maiden-run learnings; bias-to-promote idea triage (#58) and triage trigger reliability (#61); the session automation already in flight (wrapup --auto, remote control, idea queue — obot.agent PR #42); broader grant matrix and more eligible lanes as trust accrues; a /goal-style selection command. Non-negotiable boundaries: merges only via obot.agent/scripts/obot-merge with approval; the scaffold/policy carve-outs stay (goal and grant changes always wait for @jwildfire); nothing is ever deleted without approval; roadmap-first — new work found mid-run is filed, not built. What done looks like: routine roadmap increments run end-to-end unattended — idea → requirement → build → draft PR → digest — with @jwildfire reviewing at the gates rather than steering in between.

Candidate children

#18 (shipped --auto v1 — anchor/reference), #58 (ideas-triage v2), #61 (triage pipeline reliability).
Future: grant-matrix expansion, /goal command, unattended-run robustness.

Notes

Retrofitted per the #53 v2 amendment (2026-07-24): membership = the sub-issue links (small tasks may live directly under the goal, #53 policy); the --auto policy binding lives in obot.agent/goals/registry.json — this goal is not registry-listed yet, so it is display-only for --auto.

Promoted from discussion #54: https://github.com/jwildfire/obot.roadmap/discussions/54

This Issue was drafted by Claude Code using Fable 5 and reviewed by @jwildfire

Members are generated from the goal issue's sub-issue links at build time; priority is the selecting session's judgment, not list order (#53 v2). The --auto policy binding — active/paused, grant profile, repo-level backlog feeds — lives in obot.agent/goals/registry.json. Readiness labels: auto = ready for autonomous implementation, draft = needs @jwildfire steering.