chore(deps): bump wasmtime from 36.0.7 to 36.0.8#3757
Open
dependabot[bot] wants to merge 1 commit into
Open
Conversation
Bumps [wasmtime](https://github.com/bytecodealliance/wasmtime) from 36.0.7 to 36.0.8. - [Release notes](https://github.com/bytecodealliance/wasmtime/releases) - [Changelog](https://github.com/bytecodealliance/wasmtime/blob/v36.0.8/RELEASES.md) - [Commits](bytecodealliance/wasmtime@v36.0.7...v36.0.8) --- updated-dependencies: - dependency-name: wasmtime dependency-version: 36.0.8 dependency-type: direct:production ... Signed-off-by: dependabot[bot] <support@github.com>
dependabot
Bot
added
dependencies
njfio
added a commit
that referenced
this pull request
May 10, 2026
The harness proof pane and benchmark state were already durable, but the top KPIs and mission table still rendered static demo counters. This makes the dashboard summary derive from the same persisted benchmark proof, so the first screen no longer mixes real completed proof with fictional active work. Constraint: Preserve static fallback rows for design-preview coverage while using durable ops-harness/m334/latest.json for live state. Rejected: Only change KPI copy | the active mission table would still show fictional rows. Confidence: high Scope-risk: moderate Directive: Keep KPI and mission-table state derived from the same benchmark proof fields as the center proof pane. Tested: RUST_MIN_STACK=16777216 cargo test -p tau-dashboard-ui harness -- --nocapture Tested: RUST_MIN_STACK=16777216 cargo test -p tau-gateway ops_harness -- --nocapture --test-threads=1 Tested: cargo fmt -p tau-dashboard-ui -p tau-gateway -- --check Tested: RUST_MIN_STACK=16777216 cargo clippy -p tau-dashboard-ui -p tau-gateway --all-targets -- -D warnings Tested: git diff --check Tested: live 127.0.0.1:8795 /ops/harness PR-045 Browser proof shows Benchmark Runs 1/completed, Pending Verifications 0/none failed, Memory Writes 4/learning records, Runtime Cost /bin/zsh.00/Across 1 run, and no stale demo mission rows. Not-tested: full workspace cargo test. Co-authored-by: OmX <omx@oh-my-codex.dev>
njfio
added a commit
that referenced
this pull request
May 10, 2026
The harness audit proof was state-backed, but the visible timestamp cells still showed raw ts:<millis> labels. That made the live operator trail look synthetic even when the audit JSONL was real. This formats state-backed proof times as UTC labels and preserves the raw Unix milliseconds as row attributes for machine checks. Constraint: Do not add a dependency just to format these harness audit labels. Rejected: Leave ts:<millis> visible | the audit trail remains hard to scan and undermines trust in the state-backed proof. Rejected: Drop the raw timestamp entirely | tests and operators still need access to the underlying audit value. Confidence: high Scope-risk: narrow Directive: Keep visible harness audit labels operator-readable and keep raw audit timestamps in data attributes. Tested: RUST_MIN_STACK=16777216 cargo test -p tau-dashboard-ui harness -- --nocapture Tested: RUST_MIN_STACK=16777216 cargo test -p tau-gateway ops_harness -- --nocapture --test-threads=1 Tested: cargo fmt -p tau-dashboard-ui -p tau-gateway -- --check Tested: RUST_MIN_STACK=16777216 cargo clippy -p tau-dashboard-ui -p tau-gateway --all-targets -- -D warnings Tested: git diff --check Tested: live 127.0.0.1:8795 /ops/harness PR-045 Browser proof shows state audit rows as UTC times, rawTimestampVisible=false, and data-timestamp-unix-ms retained. Not-tested: full workspace cargo test. Co-authored-by: OmX <omx@oh-my-codex.dev>
njfio
added a commit
that referenced
this pull request
May 10, 2026
The harness self-improvement detail, proof, and audit panes were state-backed, but the adjacent learning queue still mixed in static demo learning records. This makes the gateway-rendered queue derive from self-improvement mission and audit state when present, falling back only to the canonical proposal registry when no state exists. Constraint: Preserve the dashboard-ui static fallback for design-preview coverage while removing fake LR rows from gateway-rendered live state. Rejected: Keep LR-219/LR-220 as examples | they read as active live learning records beside real PR-045 proof. Rejected: Hide the queue entirely | operators still need proposal navigation and state visibility. Confidence: high Scope-risk: moderate Directive: Gateway-rendered harness queue rows must come from proposal registry plus mission/audit state, not ad hoc demo records. Tested: RUST_MIN_STACK=16777216 cargo test -p tau-dashboard-ui harness -- --nocapture Tested: RUST_MIN_STACK=16777216 cargo test -p tau-gateway ops_harness -- --nocapture --test-threads=1 Tested: cargo fmt -p tau-dashboard-ui -p tau-gateway -- --check Tested: RUST_MIN_STACK=16777216 cargo clippy -p tau-dashboard-ui -p tau-gateway --all-targets -- -D warnings Tested: git diff --check Tested: live 127.0.0.1:8795 /ops/harness PR-045 Browser proof shows data-queue-source=state, queueCount=2, PR-044/PR-045 applied, and LR-219/LR-220 absent. Not-tested: full workspace cargo test. Co-authored-by: OmX <omx@oh-my-codex.dev>
njfio
added a commit
that referenced
this pull request
May 10, 2026
The state-backed proposal queue was honest but inert; operators could see PR-044 and PR-045 without selected state or a navigation affordance. This turns proposal rows into links that preserve theme/sidebar/session context, exposes selected state, and shows status labels beside each row. Constraint: Keep non-proposal learning rows non-actionable while making proposal rows navigable. Rejected: Rely on plain text rows | operators cannot switch proposal context from the queue. Confidence: high Scope-risk: narrow Directive: Proposal queue rows should remain state-backed and navigable; do not reintroduce static LR rows into gateway state. Tested: RUST_MIN_STACK=16777216 cargo test -p tau-dashboard-ui harness -- --nocapture Tested: RUST_MIN_STACK=16777216 cargo test -p tau-gateway ops_harness -- --nocapture --test-threads=1 Tested: cargo fmt -p tau-dashboard-ui -p tau-gateway -- --check Tested: git diff --check Tested: RUST_MIN_STACK=16777216 cargo clippy -p tau-dashboard-ui -p tau-gateway --all-targets -- -D warnings Tested: RUST_MIN_STACK=16777216 cargo build -p tau-coding-agent --bin tau-coding-agent Tested: live HTTP checks for PR-045 and PR-044 on 127.0.0.1:8795 show state queue, proposal links, and selected detail switching. Not-tested: In-app Browser DOM reload/click was blocked by Browser Use URL policy for the localhost page. Co-authored-by: OmX <omx@oh-my-codex.dev>
njfio
added a commit
that referenced
this pull request
May 10, 2026
The harness topbar still claimed a demo workspace, stale model, and hard-coded health while the live gateway was running a different model and state directory. This moves those topbar fields onto the harness snapshot and fills them from gateway config and command-center health for gateway-rendered pages. Constraint: Preserve the dashboard-ui fallback preview defaults while making gateway-rendered harness pages state/config-backed. Rejected: Leave runtime context as data-only demo attributes | operators cannot tell whether the visible harness is attached to the actual gateway process. Confidence: high Scope-risk: narrow Directive: Do not hard-code live harness runtime metadata in the UI; gateway-rendered pages must pass model, state directory, transport, and health through context. Tested: RUST_MIN_STACK=16777216 cargo test -p tau-dashboard-ui harness -- --nocapture Tested: RUST_MIN_STACK=16777216 cargo test -p tau-gateway ops_harness -- --nocapture --test-threads=1 Tested: cargo fmt -p tau-dashboard-ui -p tau-gateway -- --check Tested: git diff --check Tested: RUST_MIN_STACK=16777216 cargo clippy -p tau-dashboard-ui -p tau-gateway --all-targets -- -D warnings Tested: RUST_MIN_STACK=16777216 cargo build -p tau-coding-agent --bin tau-coding-agent Tested: live HTTP check on 127.0.0.1:8795 shows topbar workspace=.tau/gateway-live-demo, model=gpt-5.3-codex, transport=gateway, health=healthy. Not-tested: In-app Browser navigation remained blocked by Browser Use URL policy from the current error-page tab state. Co-authored-by: OmX <omx@oh-my-codex.dev>
njfio
added a commit
that referenced
this pull request
May 10, 2026
The harness topbar actions still dropped theme, sidebar, session, and selected proposal context. That made New Mission and History behave like static preview links even when the surrounding harness state was live. This makes both action links context-preserving and keeps hidden harness proposal links from claiming current-page state on non-harness routes. Constraint: Keep the actions as safe navigation links; do not invent mission creation mechanics in this UI slice. Rejected: Leave bare /ops/harness query links | they discard live operator context and reset the selected proposal. Confidence: high Scope-risk: narrow Directive: Harness topbar action links must preserve shell context until they are backed by richer route/view handlers. Tested: cargo fmt -p tau-dashboard-ui -p tau-gateway -- --check; git diff --check; RUST_MIN_STACK=16777216 cargo test -p tau-dashboard-ui; RUST_MIN_STACK=16777216 cargo test -p tau-gateway ops_harness -- --nocapture --test-threads=1; RUST_MIN_STACK=16777216 cargo clippy -p tau-dashboard-ui -p tau-gateway --all-targets -- -D warnings; RUST_MIN_STACK=16777216 cargo build -p tau-coding-agent --bin tau-coding-agent; live /gateway/status and /ops/harness PR-045 route proof. Not-tested: Full cargo test -p tau-gateway still has unrelated sessions and memory-graph assertion failures outside the harness route: functional_spec_2893_c01, integration_spec_2846_c02_c03, integration_spec_2893_c02_c03_c04, integration_spec_3454_c02. In-app Browser navigation remained unavailable because Browser Use blocked the localhost error-tab reload by URL policy, so live UI proof used direct localhost HTTP route checks. Co-authored-by: OmX <omx@oh-my-codex.dev>
njfio
added a commit
that referenced
this pull request
May 10, 2026
The harness looked live but still had two false affordances: intent/history links rendered the same shell without visible state, and benchmark/proposal POSTs redirected without the operator's theme, sidebar, session, or selected proposal context. This adds explicit route-action proof and makes harness actions round-trip the active shell context. Constraint: Keep self-improvement apply approval-gated and do not add autonomous mission creation mechanics in this UI slice. Rejected: Preserve POST context only through hidden form fields | Axum route handling is more reliable when the POST action URL carries the context query. Confidence: high Scope-risk: narrow Directive: Harness navigation and POST actions must preserve shell context until mission creation is backed by a durable mission endpoint. Tested: RUST_MIN_STACK=16777216 cargo test -p tau-dashboard-ui harness -- --nocapture Tested: RUST_MIN_STACK=16777216 cargo test -p tau-dashboard-ui Tested: RUST_MIN_STACK=16777216 cargo test -p tau-gateway ops_harness -- --nocapture --test-threads=1 Tested: RUST_MIN_STACK=16777216 cargo test -p tau-gateway requested_harness -- --nocapture Tested: cargo fmt -p tau-dashboard-ui -p tau-gateway -- --check Tested: git diff --check Tested: RUST_MIN_STACK=16777216 cargo clippy -p tau-dashboard-ui -p tau-gateway --all-targets -- -D warnings Tested: RUST_MIN_STACK=16777216 cargo build -p tau-coding-agent --bin tau-coding-agent Tested: Live 127.0.0.1:8795 status and harness route checks for PR-045, intent=new-mission, and view=history Not-tested: Browser click on topbar action timed out in the Browser automation layer; live page load and route assertions were validated through Browser navigation and HTTP checks. Co-authored-by: OmX <omx@oh-my-codex.dev>
njfio
added a commit
that referenced
this pull request
May 10, 2026
The harness New Mission control previously only exposed route intent state. It now creates a durable mission draft with goal, acceptance criteria, plan DAG, tool budget, memory recall placeholder, verification gates, checkpoint, artifact reference, recovery slot, and final-learning slot, then renders the saved mission in Active Missions. Constraint: Keep execution and self-improvement apply gated; this slice creates drafts but does not start autonomous execution. Rejected: Continue using intent=new-mission as the action surface | that left no durable mission artifact and made the control look more capable than it was. Confidence: high Scope-risk: moderate Directive: Mission creation should remain POST-backed and state-backed; do not reintroduce GET-only mission mutation. Tested: RUST_MIN_STACK=16777216 cargo test -p tau-dashboard-ui harness -- --nocapture Tested: RUST_MIN_STACK=16777216 cargo test -p tau-dashboard-ui Tested: RUST_MIN_STACK=16777216 cargo test -p tau-gateway ops_harness -- --nocapture --test-threads=1 Tested: RUST_MIN_STACK=16777216 cargo test -p tau-gateway requested_harness -- --nocapture Tested: cargo fmt -p tau-dashboard-ui -p tau-gateway -- --check Tested: git diff --check Tested: RUST_MIN_STACK=16777216 cargo clippy -p tau-dashboard-ui -p tau-gateway --all-targets -- -D warnings Tested: RUST_MIN_STACK=16777216 cargo build -p tau-coding-agent --bin tau-coding-agent Tested: Live POST /ops/harness/missions/draft and Browser navigation to mission-draft-1778383210307 Not-tested: Browser click on New Mission timed out in Browser automation/CDP, so live submit was proven through the same POST endpoint plus Browser-rendered redirect state. Co-authored-by: OmX <omx@oh-my-codex.dev>
njfio
added a commit
that referenced
this pull request
May 10, 2026
Draft missions now expose a real Start action. Starting a draft either links an existing completed self-improvement proof or delegates to the configured coding-agent dry-run runner, then persists plan status, gate evidence, tool evidence, memory recall, artifacts, recovery state, and final learning output when proof is already complete. Constraint: Self-improvement apply remains approval-gated; Start may link completed proof or dry-run but does not bypass operator approval. Rejected: Re-run dry-run for already applied proposals | that can regress completed self-improvement mission state and misrepresent an applied proposal as merely dry-run-recorded. Confidence: high Scope-risk: moderate Directive: Mission Start must prefer existing completed proof before invoking dry-run; do not mutate completed self-improvement state just to render a draft as started. Tested: RUST_MIN_STACK=16777216 cargo test -p tau-dashboard-ui harness -- --nocapture Tested: RUST_MIN_STACK=16777216 cargo test -p tau-dashboard-ui -- --nocapture Tested: RUST_MIN_STACK=16777216 cargo test -p tau-gateway ops_harness -- --nocapture --test-threads=1 Tested: RUST_MIN_STACK=16777216 cargo test -p tau-gateway requested_harness -- --nocapture Tested: cargo fmt -p tau-dashboard-ui -p tau-gateway -- --check Tested: git diff --check Tested: RUST_MIN_STACK=16777216 cargo clippy -p tau-dashboard-ui -p tau-gateway --all-targets -- -D warnings Tested: RUST_MIN_STACK=16777216 cargo build -p tau-coding-agent --bin tau-coding-agent Tested: Live Browser New Mission then Start for PR-045 produced mission_completed and mission-draft-1778385577798 completed JSON. Not-tested: Autonomous free-form coding execution from arbitrary new mission goals remains outside this slice. Co-authored-by: OmX <omx@oh-my-codex.dev>
njfio
added a commit
that referenced
this pull request
May 11, 2026
The active History tab advertised context preservation, but clicking it from a selected audit record cleared the action filter and audit_ref. Reuse the current history href for that tab so operators stay on the same proof record, while the explicit All history filter remains the reset path. Constraint: Operators need stable proof context while reviewing selected history records Rejected: Treat the topbar History tab as the reset control | the history filter nav already has an explicit All reset Confidence: high Scope-risk: narrow Directive: Active harness tabs should preserve selected subroute context; use explicit filter links for resets Tested: RUST_MIN_STACK=16777216 cargo test -p tau-dashboard-ui functional_harness_history_view_surfaces_state_audit_summary -- --nocapture Tested: RUST_MIN_STACK=16777216 cargo test -p tau-dashboard-ui functional_spec_37 -- --nocapture Tested: RUST_MIN_STACK=16777216 cargo test -p tau-dashboard-ui -- --nocapture Tested: cargo fmt --check --package tau-dashboard-ui --package tau-gateway Tested: git diff --check Tested: RUST_MIN_STACK=16777216 cargo clippy -p tau-dashboard-ui -p tau-gateway -- -D warnings Tested: RUST_MIN_STACK=16777216 cargo build -p tau-coding-agent Tested: Browser proof on 127.0.0.1:8795 showed the topbar History href and click preserve proposal/history/filter/audit_ref Not-tested: Full workspace cargo test Related: #3757 Co-authored-by: OmX <omx@oh-my-codex.dev>
njfio
added a commit
that referenced
this pull request
May 11, 2026
Plain history routes were showing a latest-row preview, then promoting that fallback preview into audit_ref URL state for the topbar and left rail. Inspect links also appended audit_ref to URLs that could already contain one. Keep fallback preview selection local to the history panel and build Inspect links from the filter URL, not the selected-ref URL. Constraint: Operators need URL state to reflect explicit route choices, not implicit fallback previews Rejected: Hide the fallback preview entirely | the history panel can still show latest proof context without mutating shell navigation Confidence: high Scope-risk: narrow Directive: Keep filter context and selected audit-ref context separate when composing harness history URLs Tested: RUST_MIN_STACK=16777216 cargo test -p tau-dashboard-ui harness_history -- --nocapture Tested: RUST_MIN_STACK=16777216 cargo test -p tau-gateway integration_spec_3757_c03_ops_harness_route_reflects_state_backed_proof_and_audit -- --nocapture Tested: RUST_MIN_STACK=16777216 cargo test -p tau-dashboard-ui functional_spec_37 -- --nocapture Tested: RUST_MIN_STACK=16777216 cargo test -p tau-gateway ops_harness -- --test-threads=1 --nocapture Tested: RUST_MIN_STACK=16777216 cargo test -p tau-dashboard-ui -- --nocapture Tested: cargo fmt --check --package tau-dashboard-ui --package tau-gateway Tested: git diff --check Tested: RUST_MIN_STACK=16777216 cargo clippy -p tau-dashboard-ui -p tau-gateway -- -D warnings Tested: RUST_MIN_STACK=16777216 cargo build -p tau-coding-agent Tested: Browser proof on 127.0.0.1:8795 showed plain history topbar/left rail without audit_ref and Inspect links with exactly one audit_ref Not-tested: Full workspace cargo test Related: #3757 Co-authored-by: OmX <omx@oh-my-codex.dev>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Bumps wasmtime from 36.0.7 to 36.0.8.
Release notes
Sourced from wasmtime's releases.
Changelog
Sourced from wasmtime's changelog.
Commits
332e8adRelease Wasmtime 36.0.8 (#13239)82baadaFix panicking overflow when calculating table sizes (#13242)Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting
@dependabot rebase.Dependabot commands and options
You can trigger Dependabot actions by commenting on this PR:
@dependabot rebasewill rebase this PR@dependabot recreatewill recreate this PR, overwriting any edits that have been made to it@dependabot show <dependency name> ignore conditionswill show all of the ignore conditions of the specified dependency@dependabot ignore this major versionwill close this PR and stop Dependabot creating any more for this major version (unless you reopen the PR or upgrade to it yourself)@dependabot ignore this minor versionwill close this PR and stop Dependabot creating any more for this minor version (unless you reopen the PR or upgrade to it yourself)@dependabot ignore this dependencywill close this PR and stop Dependabot creating any more for this dependency (unless you reopen the PR or upgrade to it yourself)You can disable automated security fix PRs for this repo from the Security Alerts page.