Skip to content

Releases

v1.5.3

Choose a tag to compare

@sblattj sblattj released this 08 Nov 23:25

Docs and release tooling only. The shipped src/ and dist/ are unchanged from v1.5.2.

Changed

  • On npm. npx -y opencode-goal-pro-max-complete-plugin install is now the main install command. The github: form still works and is listed as the alternative in the README, docs/install.md and the installer's help.
  • Releases publish themselves. Each v* tag runs the release gate, then publishes the verified tarball to npm through trusted publishing, with provenance. A version npm already has is skipped.

v1.5.2

Choose a tag to compare

@sblattj sblattj released this 03 Nov 03:13

Contributor-facing release. The shipped src/ and dist/ are byte-identical to v1.5.1, so upgrading changes nothing at runtime.

Changed

  • Bun 1.3.14 is the development runtime. Tests, coverage, the mutation contract, smokes, benchmarks, verify, check and bundle run under Bun. Install with bun install --frozen-lockfile; bun.lock replaces package-lock.json.
  • The npx installer stays on Node. CI keeps a Node installer (npx path) job on Node 18, the engines floor. The Node 18/20/22/24 unit-test matrix is gone.
  • Mutation contract: a mutant only counts as killed when a test really fails, and each run stops at the first failure. One old mutant never compiled; it now does, and is killed.
  • Audit: bun audit --prod --audit-level=high replaces npm audit.
  • Windows is not supported. The Windows CI job is gone and the docs say so.

Full details in CHANGELOG.md.

v1.5.1

Choose a tag to compare

@sblattj sblattj released this 01 Nov 22:04

Use this instead of v1.5.0. The v1.5.0 bundles were built from a stale install, so dist/ carried zod 4.4.3 instead of the lockfile's 4.6.7. That undid the 1.4.1 security update (GHSA-mrhx-c7xc-m3w5, high).

Rebuilt from a clean npm ci with the pinned Bun 1.3.14. No other change. See the v1.5.0 notes for forced compaction, the feature these releases add.

v1.5.0

Choose a tag to compare

@sblattj sblattj released this 01 Nov 21:48

Use v1.5.1 instead. This build carried zod 4.4.3, undoing the 1.4.1 security update.

Goal sessions now compact themselves. OpenCode never auto-compacts a model that reports limit.context: 0, so a long goal on such a model kept growing (638,078 tokens in the case that prompted this).

Added: forced compaction, on by default at 200,000 tokens

  • --compact-at <n> and compactAtTokens set the threshold; 0 turns it off. It applies on every model, so a 1M-window model also compacts at 200k unless you raise it.
  • Checked after every finished step, so one long turn is caught mid-turn. The idle check is the backstop.
  • Never two compactions for one crossing. On a model with a real limit.context, the host's own check gets one step first.
  • A failed compaction does not block the goal: the plugin deletes the failed compaction message, because the host would otherwise re-run it on later prompts.
  • Bounded: 3 failed requests stop forcing until the next compaction, and forcing is held when two compactions in a row leave the context over the threshold.
  • Goal sessions only. Forcing never pauses or stops a goal.

Added: npm run smoke:forced-compaction, which runs the feature against a real OpenCode host with a mock model.

Checked against OpenCode 1.19.0 (running) and the v1.19.5 source. Full details in CHANGELOG.md and docs/reference.md.

v1.4.1

Choose a tag to compare

@sblattj sblattj released this 27 Oct 02:12
  • Fixed: context fullness no longer stops a goal or tells it to wrap up. On 1.4.0 a goal with 7.2M of 100M tokens spent and 18m of 8h elapsed paused with "budget wrap-up requested" because peak context reached 80% of the model's 128,000-token window. OpenCode compacts the session on its own when context fills, so context is not a budget. The context window reached stop, the context branch of the 80% wrap-up, the "context token(s) remaining" warning and the context_remaining line in the continuation prompt are gone. Token spend (maxTokens) and the clock (maxDurationMs) keep both their 80% wrap-up and their hard stop.
    • A wrap-up names its budget. The stop reason is token budget wrap-up requested or time budget wrap-up requested, and the prompt opens "Token budget limit near" or "Time budget limit near".
    • Status and panel show current context. /goal status and the sidebar show the latest message's context, which drops on compaction, instead of a lifetime high-water mark. The status label is Context: and says where the window size came from: configured (--context-window) or auto (host catalog: <provider>/<model>).
    • The auto-detected window prefers limit.input when the model reports one, as OpenCode's own overflow check does (checked against OpenCode v1.19.4 session/overflow.ts).
    • contextWindowTokens and --context-window are display only. Still accepted; they no longer stop or nudge anything.
  • Security: bundled zod updated from 4.4.3 to 4.6.7 (GHSA-mrhx-c7xc-m3w5, high).

v1.4.0

Choose a tag to compare

@sblattj sblattj released this 08 Oct 19:30
  • Changed: a goal stops only when you stop it, when it finishes, when a budget runs out, or when the model is genuinely hard-blocked. Every other automatic stop now records what happened and keeps going. docs/reference.md has a single "When a goal stops" section.
    • Compaction is not a human. OpenCode writes a user-role message for each compaction (a compaction part, then a synthetic "Continue if you have next steps" text). The plugin counted those as human turns, so with noInterruptOnUserMessage: false (the default before 1.2.0) every compaction paused the goal with "user intervention". isHostSyntheticUserMessage now excludes them.
    • Your messages steer, never pause. noInterruptOnUserMessage is deprecated: accepted, ignored, always on. No more "Stopped: user intervention".
    • Turn-ending host events end the turn, not the goal. Permission rejections, aborted turns (Escape), provider errors and attachment-resolution errors are recorded and the goal continues after the cooldown. Repeated provider errors back off 5 s, 10 s, 20 s … up to 5 minutes.
    • Loop brakes nudge instead of stopping. Stalled compactions (no longer aborting the session), low-progress turns and tool-free turns add a "Stalled:" re-plan instruction to the next continuation, and the cooldown backs off exponentially (minDelayMs × 2^strikes, up to 64×). Format-validation and auto-continue failures keep retrying with backoff.
    • New option promptRetryWakeMs (default 5000). A failed auto-continue prompt starts no turn, so no idle follows it; the plugin now retries on its own after this delay (or the backed-off cooldown, if longer).
    • A rejected completion audit keeps the goal active and sends the auditor's reason with the next continuation.
    • Planning-only agents defer, not stop. Under Plan the goal's clock freezes; it resumes under the next executing agent, with that agent's context.
    • Restart keeps a running goal running. A goal that was active before a plugin restart continues at the next idle; downtime does not count against its duration budget. A newer pause or halt in the lifecycle ledger still wins over a stale snapshot, and a goal rebuilt from the ledger alone still loads paused.
  • New: /goal halt, the user's stop (an alias of /goal pause, resumable with /goal resume). /goal status suggests halt/resume/clear while a goal runs.
  • Removed: the goal_pause tool (13 tools now). update_goal and the canonical update path refuse status: "paused" with pause_user_only. Pausing is user-only.
  • Changed: blocking is deterred. goal_block and [goal:blocked] still work with a concrete blocker, but every prompt surface, tool description, the shipped skill and the docs now say that blocking stops the whole goal and is only for input that only you can give; a mid-goal message from you is steering, not a reason to block.
  • Unchanged: budget ceilings, /goal pause and /goal clear, completion, queued/backgrounded goals, and the two-process persistence fences (terminal persistence failed, continuation claim persistence failed).

v1.3.0

Choose a tag to compare

@sblattj sblattj released this 05 Oct 01:26
  • Fixed: "Goal plugin is locked by another opencode process" no longer pins a session to whichever process touched it first. Before this release, a process took a session's lease the first time any hook saw that session, even an ordinary chat message with no goal, and kept it until the process exited. A second OpenCode process opening the same session was passive for good. Now a process holds a session's lease only while it is driving it. That means an active or queued goal, a /goal turn, a continuation, a prompt, a session load, or a hook or tool call still running for it. After the session has been idle for idleLeaseReleaseMs (new option, default 3000; 0 restores the old keep-until-exit behaviour), the owner persists, drops its in-memory copy, and releases the lease. Paused, stopped, and finished goals do not hold it. Goals in different sessions already ran concurrently, one shard per session. Two processes can now also hand a single session back and forth. A session with a running goal stays exclusive to its owner, and forking (opencode --continue --fork) remains the way to run two goals at once from one conversation.
    • New: lease heartbeat and stale reclaim. Each claim records heartbeatMs and its mtime is refreshed every leaseHeartbeatMs (default 15000). A heartbeating claim older than max(leaseStaleAfterMs, 4 × heartbeat) (leaseStaleAfterMs defaults to 120000) is reclaimable whatever its pid or host. A crashed process, a reused pid, or a claim left by another machine on a shared filesystem no longer pins the session. Claims from older releases, which carry no heartbeatMs, keep the earlier pid/host rule, and also become reclaimable after 24 hours when they come from another host. The claim layout and protocol (claims-v2) are unchanged, so 1.2.x and 1.3.0 processes interoperate. A 1.2.x owner still never releases early, though.
    • New: writes are fenced after a lost lease. Before each persist, ledger append, and goal-driving prompt, the owner checks that its claim is still there. If it was reclaimed or removed, the process stops touching that session, logs a warning, and unloads it, so two processes never write one shard. Behaviour change: a host that sleeps for longer than the stale window (about 2 minutes by default) can wake to find another process has taken over its session. It then stands down instead of writing.
    • Changed: the passive-mode message says what to do. It now explains that an idle owner releases the session within a few seconds, so you should retry, and that opencode --continue --fork is the way to work in parallel.
    • New: npm run smoke:multi-process. It runs two real opencode serve processes on one project. It asserts that goals in distinct sessions run at once, that a session left idle by process A is taken over by process B, and that a session with a running goal stays with its owner. Like the smoke:todo-* rungs, it needs the opencode binary and is not part of release:check. Against 1.2.0 the takeover arm fails, and against 1.3.0 all three arms pass.
    • Types and docs. index.d.ts adds the three options to GoalPluginOptions. docs/reference.md (state and persistence, options table, troubleshooting) and docs/compatibility.md describe idle release, the heartbeat and stale rule, and the sleep caveat.

Install: opencode-goal-pro-max-complete-plugin@github:sblattj/opencode-goal-pro-max-complete-plugin#v1.3.0 (see docs/reference.md).

v1.2.0

Choose a tag to compare

@sblattj sblattj released this 03 Oct 19:25
  • Changed: a human message no longer stops a running goal. noInterruptOnUserMessage now defaults to true. When you type a message while a goal is running, the goal keeps going: OpenCode answers your message, and the next auto-continue picks up with it as the latest instruction. Before this change the goal paused with stopReason: "user intervention" and waited for /goal resume. /goal pause and /goal stop are now how you halt a goal. To get the old behavior back, set noInterruptOnUserMessage: false. An explicit false is the only value that turns pausing back on, so a non-boolean such as "yes" now also counts as true. The index.d.ts JSDoc, docs/reference.md, and the shipped skill describe the new default.

Install: opencode-goal-pro-max-complete-plugin@github:sblattj/opencode-goal-pro-max-complete-plugin#v1.2.0 (see docs/reference.md).

v1.1.0

Choose a tag to compare

@sblattj sblattj released this 07 Oct 02:56
  • New: OpenCode 2 hosts are supported, alongside OpenCode 1. The package's default export is now the dual-method object { id, server, setup }. OpenCode 1.x keeps calling server(input, options) exactly as in every earlier release, and OpenCode 2.x calls the new setup(context) entrypoint, which rebuilds the same plugin — the goal tools, the event-driven continuation engine, persistence and leases, and the sidebar status — on the v2 host API and returns the dispose finalizer the v2 host awaits at shutdown. Both hosts load the same bundled module through the same package exports (. and ./server; both host generations resolve ./server), the hosts self-select by which method they call, and an install upgraded from 1.0.1 needs no config change and keeps its goals, persisted state and wire identifiers. The v1 hook surface, the 14 tools, the /goal command semantics on v1 hosts, the persisted-state format, and the metadata.goal payload (v: 3) are all unchanged. engines.opencode widens from >=1.17.15 <2 to >=1.17.15 <3: OpenCode 1.x enforces the field at load time, while OpenCode 2 ignores it (verified live), so the widened bound gates 1.x loads and documents the supported span rather than restricting 2.x.
    • What OpenCode 2 does not provide, and how the plugin fails soft. v2 has no equivalent for a handful of host surfaces, and every affected call site is individually guarded, so each gap degrades that one surface and nothing else — none of them fails the plugin or interrupts a running goal. No command.execute.before equivalent: v2 has no pre-execution command interception and no noReply-style suppression of the command turn's assistant echo. A typed /goal … still reaches the plugin as prompt text — the v2 adapter re-runs the v1 command handler for its state side effects (set, pause, resume, stop, clear, ledger, persistence) inside the prompt hook, on by default and disableable with the v2-side v2CommandTextBridge plugin option — but the deterministic framed reply cannot be injected into the turn, so the goal_* agent tools remain the surface whose result reaches the model verbatim, with an exact tool equivalent for every state change /goal performs. No TUI toast or app.log routes: the lifecycle and audit advisories that route through client.tui.showToast and client.app.log fail soft (the dispatch is advisory and swallows the failure), and notices still reach the user through the session title and tool results. No session.todo, session.children, session.status or config.providers routes: the <existing_todos> offer is skipped, noContinueWhileChildrenActive fails open (continuation proceeds when the host cannot report children), and the auto-detected context-window ceiling stays off unless contextWindowTokens is configured explicitly. The full caveat list, and what "supported" is verified against, is in docs/compatibility.md, rewritten this release from "not supported, and not yet tested" to supported-with-caveats.
    • The published type declarations carry the new entrypoint. index.d.ts's default export gains the optional setup?: (context: unknown) => Promise<() => Promise<void>> beside server, with the v2 context kept loose (unknown) like the hook payloads; GoalPluginHooks, GoalPluginOptions, and both payload-schema version strings (GoalSidebarStatus.v, GOAL_PANEL_PAYLOAD_VERSION) are unchanged. npm run type:check's consumer fixture and npm run smoke:packed-host now require all three export keys — id, server and a callable setup — so a build that drops the v2 entrypoint fails the gate instead of a v2 host.
    • Docs follow the widened claim. README.md and the docs/reference.md compatibility table now state >=1.17.15 <3 with the 1.x-and-2.x span; docs/providers.md marks its matrix as 1.x evidence records and points 2.x rows at the compatibility policy; the install specs throughout docs/reference.md and examples/ pin v1.1.0. The five-step "what a supported v2 claim would require" checklist in docs/compatibility.md is retired — its conditions are the release's own bar now.

Install: opencode-goal-pro-max-complete-plugin@github:sblattj/opencode-goal-pro-max-complete-plugin#v1.1.0 (see docs/reference.md).

v1.0.1

Choose a tag to compare

@sblattj sblattj released this 07 Oct 02:56
  • New: the session's native Todo list is drawn from the goal plan. While a plan exists the plugin rewrites the arguments of the model's own todowrite call in tool.execute.before, so the host performs exactly one write and the checklist every client renders is the plan: one row per action in plan order, content a3 · rebuild dist (the action id, ·, and the title bounded to 120 characters), and a plan past 20 actions ending in one counted +N more actions — /goal status row. The status map exists to stop the native [✓] asserting something the plugin's own evidence gate refuses — a done action becomes completed only once it is verified, a done action without a passing verdict mirrors as in_progress with — needs claim/evidence/verdict, and a blocked action mirrors as in_progress with — BLOCKED: <reason>, which is what the host's own todo contract asks for. Rows the model wrote itself are kept, below the plan rows, at most 10 of them and each bounded to 120 characters; nothing of the model's is silently deleted, and the tool result names how many were kept and how many the cap trimmed. todowrite({todos: []}) is the refresh idiom and it is a redraw, never a delete: with a live plan it re-emits the projection and leaves the kept rows alone, with no live plan it re-emits the last mirrored rows or the terminal snapshot taken when the goal ended, and only in a session that never mirrored anything does an empty call reach the host as an empty list. A stale mirror is nudged at most three times per goal run — Todo panel is stale — call todowrite({todos: []}) once; the plan is copied into it for you. — shared across the auto-continue block and goal_action_update's result, and /goal resume refunds the budget in full. /goal stop, /goal clear and a genuine completion each hand the list back in one line, because the mirror is one-way and the rows outlive the goal.
    • New option mirrorTodos: "plan" (default) or "off". "off" restores 1.0.0 exactly: all three todowrite hooks return before touching anything, every prompt surface that mentions the mirror is silent, and goal_plan_set's, goal_action_update's and todowrite's descriptions are the stock bytes — the <existing_todos> offer below is skipped along with the host read that produces it, so an "off" instance makes no session.todo call at all. The option is static and instance-wide by construction, so it cannot introduce cross-session skew.
    • There is no "adopt" mode. A session that already has native todo rows when /goal set runs gets them described back to the model once, in an <existing_todos> block on the goal's first continuation, with the instruction to either record them as the plan through goal_plan_set — rewritten as falsifiable claims — or ignore them. The read is best-effort and read-only: nothing is adopted, goal.mirror.extra starts empty, and a failing or absent host call is logged and skipped rather than blocking /goal set.
    • The Goal panel becomes an exception list while the mirror is on, because the Todo section is already showing the plan: unverified done actions first, then blocked, then in_progress, each group in plan order, and pending actions are not repeated. A done action without a verdict gains the same — needs claim/evidence/verdict suffix on its panel line that its mirrored row carries, and the progress line gains · todo mirror fresh (5), · todo list stale, or · mirror drift (7≠5) when the panel's own live read of the host's list disagrees with the payload. Under the exception list +N more counts only the exception rows the 12-row cap dropped, over the published actions rather than the whole plan, so a server-capped plan cannot promise rows the payload does not carry. A payload with no plan.mirror, or one whose state is off, renders exactly as it did in 1.0.0.
    • metadata.goal payload is now v: 3, adding two keys: plan.mirror and plan.actions[].verified. plan.mirror carries state ("fresh", "stale" or "off"), rows and extra (how many rows the last mirrored write held and how many of those were the model's own), and at (when that write landed, 0 for never). verified is the server's own CEV verdict on each published action — done and a claim and evidence and verdict: "pass" — and it exists because a consumer cannot re-derive it: claim and evidence are goal-private prose that never go on the wire, and goal_plan_set accepts a done action with a passing verdict and no ledger (only goal_action_update enforces the gate), so status plus verdict renders exactly that action as finished. The Goal panel now reads the flag rather than inferring it, which is what puts such an action back in the exception list with its — needs claim/evidence/verdict suffix, agreeing with the mirrored todo row and the progress count that already named it unverified. Every v2 key is still written and both keys are purely additive. Skew warning, unchanged in shape from the v1→v2 one: the two halves are registered separately (opencode.json and tui.json), and the panel never reads v at all, so a 0.11.0 panel against a v3 payload simply lists every action and shows no mirror suffix, and the current panel against a v2 payload does the same. Upgrade both halves together.
    • Two new hooks: tool.execute.after and tool.definition. The freshness stamp lives in the after hook because only that hook proves the host actually wrote the list — between the two hooks the host decodes the arguments and asks for permission, and either can abort the call, so a mirror stamped fresh for a write that never landed would suppress the very nudge that repairs it. tool.definition appends a static paragraph to the builtin todowrite description telling the model what the tool now does in a goal-bearing session; it is never gated on goal state, because the hook carries no session id. scripts/verify.mjs requires both, so the hook surface it checks is now eleven names. A host that does not know tool.definition simply never fires it.
    • The todowrite description suffix is stamped at most once per object. The tool.definition hook returns early when the description it was handed already carries the mirror paragraph, so a host that re-triggers the hook against the same output object — or a second registration of this plugin in one directory — cannot double it. OpenCode 1.18.29 rebuilds the description object on every ToolRegistry.tools() call, so this is defensive rather than a fix for an observed host; it is pinned by a unit that stamps one object twice and by a mutation anchor that deletes the guard.
    • Seven hazards are documented rather than fixed, in docs/compatibility.md: the suffixed todowrite description reaching goal-less sessions in the same project, todowrite: "ask" turning every refresh into a permission modal, hosts where todowrite is denied (the mirror never becomes fresh, so the panel reads stale for the whole run and the nudge self-suppresses), another plugin's tool.execute.before overwriting the mirror with no way to detect it, the host's second and hookless todowrite implementation, /undo not rewinding todos, and one deliberate departure from the host's one-in_progress convention. The same document records the transcript bubbles — which the mirror reaches through the args rewrite rather than through a plugin slot, so on 1.18.29 they show the mirrored rows and npm run smoke:todo-safety re-measures that every run — and that priority is wire-only in the TUI.
    • The published type declarations carry the whole surface. index.d.ts: GoalSidebarStatus.v becomes 3, plan.mirror ({ state: "fresh" | "stale" | "off"; rows; extra; at }) and verified: boolean on every published action join GoalSidebarStatus, mirrorTodos joins GoalPluginOptions, and tool.execute.after and tool.definition join GoalPluginHooks. tui.d.ts: GoalPanelModel gains an optional mirror (state, rows, extra, liveTodoCount, drift), and goalPanelModel gains a second parameter — goalPanelModel(raw, options?: { liveTodoCount?: number }). That parameter is optional, so a 1.0.0 call site that passes only the payload still typechecks and simply renders no drift comparison. Reading a payload is likewise unchanged; only code that constructs a GoalSidebarStatus literal has to grow the three new required fields, since that type describes what the server writes. One deliberate asymmetry between the two files: the panel's mirror.state is declared string rather than the server's three-value union, because the panel forwards whatever state the payload carried — a newer server can invent one — and it stands the progress suffix and the drift comparison down for a state it does not recognise, while still treating the mirror as active. Both files are typechecked as a consumer sees them, under NodeNext and Bundler resolution, by npm run type:check.
    • The packed type contract exercises the 1.0.1 surface, not just the 1.0.0 one. The npm run type:check fixture now sets mirrorTodos: "plan" in its satisfies GoalPluginOptions object with a @ts-expect-error negative case for an invalid value, calls tool.execute.after and tool.definition beside tool.execute.before, reads plan.mirror.state and plan.mirror.rows off a GoalSidebarStatus, and calls goalPanelModel both with and without its new options argument — so a declaration that stops compiling for a consumer fails the gate instead of waiting for the next reader.
  • New: two real-host smokes and a benchmark for the mirror. npm run smoke:todo-mirror boots a real opencode serve against a mock model and asserts the plan is redrawn over a divergent list, with two controls — the same run under mirrorTodos: "off", and a session wi...
Read more