Expose plugin approval action metadata by Guardiola31337 · Pull Request #82431 · openclaw/openclaw

Guardiola31337 · 2026-05-16T04:06:11Z

Review context

Concrete plugin: Guardiola31337/openclaw-agentkit. The intent is to keep World ID / AgentKit out of OpenClaw core while letting external HITL plugins use the existing approval flow end-to-end.

This PR's role: approval action metadata and no-route pending surface; it does not resolve verified approvals by itself.

Bundle:

#82431 exposes approval action metadata/UI rendering.
#82434 adds verified approval resolution with plugin ownership checks.
#82471 adds chat/status metadata and idempotency for durable approval updates.

sequenceDiagram
  participant Agent
  participant Gateway
  participant Plugin as HITL plugin
  participant UI as TUI/Web UI
  participant Verifier as External verifier

  Agent->>Gateway: protected tool call
  Gateway->>Plugin: approval hook/request
  Gateway->>UI: approval card + plugin actions
  UI->>Plugin: user chooses verify once / trust session / deny
  Plugin->>Verifier: external proof flow
  Verifier-->>Plugin: verified result
  Plugin->>Gateway: resolve verified approval
  Gateway->>Agent: allow/deny blocked tool
  Plugin->>UI: status/update messages with idempotency

flowchart LR
  A["Approval action surface<br/>#82431"] --> B["Verified resolver<br/>#82434"]
  B --> C["Status/chat idempotency<br/>#82471"]
  C --> D["Usable external HITL plugin flow"]

Trust boundary to review: should a plugin that already owns a pending approval be allowed to complete that specific approval after external verification, through a typed SDK/Gateway path with admin scope and ownership checks?

Summary:

Add plugin approval action templates to the before_tool_call approval result and plugin.approval.request schema.
Expand templates with the server-generated approval id, store the resulting action commands, and carry them through reply payloads and approval view models.
Add keepPendingWithoutRoute so external HITL plugins can keep plugin approvals pending when no core approval route/client is active.
Keep native decision-button channel adapters compatible with command-only approval actions.
Update the generated Swift protocol model and plugin SDK API baseline for the intentional additive surface change.

Verification:

pnpm protocol:check
pnpm plugin-sdk:api:check
pnpm tsgo:extensions
pnpm build:plugin-sdk:strict-smoke
node scripts/run-oxlint.mjs --tsconfig config/tsconfig/oxlint.extensions.json extensions/discord/src/approval-handler.runtime.ts extensions/matrix/src/approval-handler.runtime.ts extensions/qqbot/src/bridge/approval/handler-runtime.ts extensions/telegram/src/approval-handler.runtime.ts
node scripts/run-vitest.mjs src/gateway/server-methods/plugin-approval.test.ts src/infra/approval-view-model.test.ts src/infra/exec-approval-reply.test.ts src/plugin-sdk/approval-renderers.test.ts extensions/discord/src/approval-handler.runtime.test.ts extensions/matrix/src/approval-handler.runtime.test.ts extensions/qqbot/src/bridge/approval/handler-runtime.test.ts extensions/qqbot/src/engine/approval/index.test.ts src/agents/pi-tools.before-tool-call.e2e.test.ts src/plugins/hooks.before-tool-call.test.ts
node scripts/run-vitest.mjs src/plugins/contracts/host-hooks.contract.test.ts
git diff --check origin/main...HEAD

Real behavior proof

Behavior addressed: external plugins can attach approval-specific command actions to plugin approval requests and keep those requests pending while their own HITL command flow is responsible for resolution.

Real environment tested: local OpenClaw checkout on macOS, rebased onto current origin/main on May 16, 2026.

Exact steps or command run after this patch: ran the protocol generator/check, plugin SDK API baseline check, extension typecheck, plugin SDK strict declaration smoke, targeted extension oxlint, focused approval/gateway/plugin Vitest coverage, the host hook contract test, and whitespace diff check.

Evidence after fix:

$ pnpm protocol:check
wrote dist/protocol.schema.json
wrote apps/shared/OpenClawKit/Sources/OpenClawProtocol/GatewayModels.swift

$ pnpm plugin-sdk:api:check
OK docs/.generated/plugin-sdk-api-baseline.sha256

$ pnpm tsgo:extensions
(exit 0)

$ pnpm build:plugin-sdk:strict-smoke
(exit 0)

$ node scripts/run-oxlint.mjs --tsconfig config/tsconfig/oxlint.extensions.json extensions/discord/src/approval-handler.runtime.ts extensions/matrix/src/approval-handler.runtime.ts extensions/qqbot/src/bridge/approval/handler-runtime.ts extensions/telegram/src/approval-handler.runtime.ts
Found 0 warnings and 0 errors.

$ node scripts/run-vitest.mjs src/gateway/server-methods/plugin-approval.test.ts src/infra/approval-view-model.test.ts src/infra/exec-approval-reply.test.ts src/plugin-sdk/approval-renderers.test.ts extensions/discord/src/approval-handler.runtime.test.ts extensions/matrix/src/approval-handler.runtime.test.ts extensions/qqbot/src/bridge/approval/handler-runtime.test.ts extensions/qqbot/src/engine/approval/index.test.ts src/agents/pi-tools.before-tool-call.e2e.test.ts src/plugins/hooks.before-tool-call.test.ts
Test Files  10 passed (10)
Tests  122 passed (122)

$ node scripts/run-vitest.mjs src/plugins/contracts/host-hooks.contract.test.ts
Test Files  1 passed (1)
Tests  50 passed (50)

$ git diff --check origin/main...HEAD
(no output)

Observed result after fix: the protocol and SDK baselines reflect the new additive action metadata, representative adapters preserve command actions, approval view/reply rendering carries those commands, and the before_tool_call host hook contract remains compatible.

What was not tested: full release suite, live World ID callback flow, and the follow-up verified plugin-owned resolution API that is split into #82434.

Refs #82336

External plugin integration proof

Behavior addressed: the external ClawHub AgentKit plugin can consume the combined temporary OpenClaw host API branch that includes this PR's approval actions and no-route pending behavior.

Real environment tested: GitHub Actions in Guardiola31337/openclaw-agentkit on May 16, 2026, checking out Guardiola31337/openclaw@agentkit/external-plugin-host-apis and linking it into the external plugin package.

Exact steps or command run after this patch: the remote workflow checked out the temporary OpenClaw host branch, installed with frozen lockfiles, ran pnpm build gatewayWatch, ran pnpm build:plugin-sdk:strict-smoke, linked the host checkout into the AgentKit plugin, then ran pnpm test:hitl and pnpm pack:check.

Evidence after fix:

$ gh run watch 25972624778 --repo Guardiola31337/openclaw-agentkit --exit-status
✓ main CI · 25972624778
✓ build in 1m46s
  ✓ Run pnpm build gatewayWatch
  ✓ Run pnpm build:plugin-sdk:strict-smoke
  ✓ Run pnpm dev:link-openclaw .openclaw-source
  ✓ Run pnpm test:hitl
  ✓ Run pnpm pack:check

Observed result after fix: the external plugin builds against the split OpenClaw host APIs, the plugin SDK declarations expose the needed approval/action surface, the AgentKit HITL proof passes, and the package dry-run includes the expected ClawHub plugin artifact.

What was not tested: production World ID credential verification. This remote proof exercises the integrated temporary host branch because the external plugin requires this PR together with #82434 and #82471 for the full HITL path.

clawsweeper · 2026-05-16T04:09:49Z

Codex review: needs real behavior proof before merge. Reviewed May 27, 2026, 1:37 AM ET / 05:37 UTC.

Summary
The branch adds plugin approval action templates and keepPendingWithoutRoute through the before-tool-call hook, Gateway protocol, plugin approval storage/view models, channel/UI rendering, generated protocol/API baselines, and docs.

PR surface: Source +834, Tests +857, Docs +100, Generated 0, Other +8. Total +1799 across 43 files.

Reproducibility: not applicable. this is a feature/API PR rather than a reported current-main bug. The review is based on the diff, current-main source comparison, PR discussion, and supplied validation logs.

Review metrics: 1 noteworthy metric.

Public approval request fields: 2 added. actions and keepPendingWithoutRoute change the plugin/Gateway approval contract and generated protocol/API baselines.

Merge readiness
Overall: 🦪 silver shellfish
Proof: 🦪 silver shellfish
Patch quality: 🐚 platinum hermit
Result: blocked until real behavior proof from a real setup is added.

Overall follows the weaker of proof and patch quality, so missing proof can cap an otherwise strong patch.

Rank-up moves:

Add real behavior proof showing plugin command actions and decision actions in the Control UI or a live/native approval channel.
Add the AgentKit use-case docs requested by the maintainer so the API shape can be judged against concrete workflows.

Proof guidance:
Needs real behavior proof before merge: The PR supplies useful local test output and external plugin CI output, but not after-fix real UI/live-channel proof of the changed approval action behavior; screenshots, recordings, terminal/live output, linked artifacts, or redacted logs should be added with private data redacted, then updating the PR body should trigger re-review or a maintainer can request @clawsweeper re-review. After adding proof, update the PR body; ClawSweeper should re-review automatically. If it does not, the PR author or someone with repository write access can comment @clawsweeper re-review.

Mantis proof suggestion
A short Telegram Desktop recording would materially prove the visible command-text versus decision-button behavior that unit tests cannot show. A maintainer can ask Mantis to capture proof by posting a new PR comment that starts with the OpenClaw Mantis account mention, followed by:

telegram desktop proof: verify plugin approval command actions render as visible command text while decision actions use canonical /approve callbacks.

Risk before merge

This adds two public approval API/protocol fields, so maintainers need to accept the API shape and upgrade contract before it becomes a plugin SDK surface.
Approval action rendering now crosses Control UI and several native channel approval clients; green unit tests do not prove visible chat/UI behavior or callback routing in a real transport.
keepPendingWithoutRoute intentionally keeps plugin approvals pending without an active route, which depends on the broader linked resolver/status PRs and could otherwise leave users waiting until timeout.

Maintainer options:

Hold for API decision and real proof (recommended)
Require maintainer approval of the new plugin approval contract plus a screenshot, recording, terminal/live output, or redacted logs showing the changed approval action flow before merge.
Land as an accepted host API expansion
Maintainers may accept the protocol and approval-boundary risk based on the current tests and PR proof if they explicitly decide this is the right generic OpenClaw seam.
Pause until the API bundle is reviewed together
If the three linked AgentKit host API PRs need one coherent design review, pause this PR until the resolver and durable status surfaces are accepted or reshaped together.

Next step before merge
This needs maintainer API/security-boundary review and contributor-supplied real behavior proof, not an automated repair PR.

Security
Cleared: No concrete supply-chain, dependency, secret-handling, or obvious approval bypass defect was found in the diff; the approval trust-boundary expansion remains a maintainer merge-risk decision.

Review details

Best possible solution:

Land only after maintainers accept the plugin approval API/security boundary, the PR documents the concrete AgentKit use cases, and real proof shows command actions and decision actions behaving correctly in at least the Control UI or a live/native approval channel.

Do we have a high-confidence way to reproduce the issue?

Not applicable; this is a feature/API PR rather than a reported current-main bug. The review is based on the diff, current-main source comparison, PR discussion, and supplied validation logs.

Is this the best way to solve the issue?

Unclear. The split is narrower than bundling AgentKit into core and matches the plugin-boundary direction, but the API/security boundary and use-case documentation still need maintainer acceptance before this is clearly the best permanent solution.

AGENTS.md: found and applied where relevant.

Codex review notes: model gpt-5.5, reasoning high; reviewed against 95c8fc967876.

Label changes

Label justifications:

P2: This is a normal-priority plugin/Gateway API feature with broad but bounded approval-surface impact.
merge-risk: 🚨 compatibility: The PR changes public plugin hook, Gateway protocol, Swift model, docs, and plugin SDK baseline surfaces that external plugins may consume.
merge-risk: 🚨 message-delivery: The PR changes how approval actions render and route across Control UI and multiple native channel approval clients.
merge-risk: 🚨 security-boundary: The change affects the human approval flow and allows plugin-owned command actions plus no-route pending behavior in a sensitive approval path.
rating: 🦪 silver shellfish: Overall readiness is 🦪 silver shellfish; proof is 🦪 silver shellfish and patch quality is 🐚 platinum hermit.
status: 📣 needs proof: The PR needs real behavior proof before ClawSweeper can clear the contributor ask. Needs real behavior proof before merge: The PR supplies useful local test output and external plugin CI output, but not after-fix real UI/live-channel proof of the changed approval action behavior; screenshots, recordings, terminal/live output, linked artifacts, or redacted logs should be added with private data redacted, then updating the PR body should trigger re-review or a maintainer can request @clawsweeper re-review. After adding proof, update the PR body; ClawSweeper should re-review automatically. If it does not, the PR author or someone with repository write access can comment @clawsweeper re-review.
mantis: telegram-visible-proof: Mantis should capture Telegram visible proof. The PR changes visible Telegram approval behavior by keeping plugin command actions as text while decision actions use canonical approval callbacks.

Evidence reviewed

PR surface:

Source +834, Tests +857, Docs +100, Generated 0, Other +8. Total +1799 across 43 files.

View PR surface stats

Area	Files	Added	Removed	Net
Source	21	950	116	+834
Tests	18	861	4	+857
Docs	2	106	6	+100
Config	0	0	0	0
Generated	1	2	2	0
Other	1	9	1	+8
Total	43	1928	129	+1799

What I checked:

Repository policy applied: Root and scoped AGENTS.md files identify plugin SDK, Gateway protocol, approval routing, and Telegram approval behavior as compatibility/proof-sensitive review surfaces. (AGENTS.md:23, 95c8fc967876)
Public approval API expansion: The PR adds actions and keepPendingWithoutRoute to the Gateway plugin approval request schema, making this a public protocol/plugin API change rather than local-only implementation work. (src/gateway/protocol/schema/plugin-approvals.ts:40, f09bc5a65652)
Action validation and expansion: The PR introduces typed decision/command approval action templates, bounded label/template sizes, validation that command actions do not carry decisions, and server-side expansion of {id} into the generated approval id. (src/infra/plugin-approvals.ts:65, f09bc5a65652)
No-route pending behavior: The Gateway handler now skips immediate no-approval-route expiration when keepPendingWithoutRoute is true, so existing no-route failure behavior changes for plugins that opt into this new mode. (src/gateway/server-methods/approval-shared.ts:325, f09bc5a65652)
Control UI rendering path: The Control UI now renders plugin command actions as command rows while preserving decision buttons for Gateway-resolvable approval decisions. (ui/src/ui/views/exec-approval.ts:160, f09bc5a65652)
Current main check: Current origin/main does not contain keepPendingWithoutRoute or the new plugin approval action template types, so the central behavior is not already implemented on main. (95c8fc967876)

Likely related people:

steipete: Current-main blame for the plugin approval request payload and Gateway handler points to recent approval/Gateway work in commit d2711c9. (role: recent area contributor; confidence: high; commits: d2711c900d7b; files: src/infra/plugin-approvals.ts, src/gateway/server-methods/plugin-approval.ts, src/infra/exec-approval-forwarder.ts)
joshavant: The async before_tool_call approval path that this PR extends was added in commit 6ade9c4. (role: introduced related hook behavior; confidence: medium; commits: 6ade9c474cf1; files: src/agents/pi-tools.before-tool-call.ts, src/plugins/hook-types.ts)
vincentkoc: Commit 0e47ce5 restored approval queue targeting and plugin id prefixes in the same Gateway plugin approval area. (role: recent approval routing contributor; confidence: medium; commits: 0e47ce58bc1b; files: src/gateway/server-methods/plugin-approval.ts)
gumadeiras: Commit 6484b41 added pending approval replay behavior in the Gateway plugin approval path. (role: adjacent approval lifecycle contributor; confidence: medium; commits: 6484b41eb9e4; files: src/gateway/server-methods/plugin-approval.ts)
kevinslin: The live discussion includes a maintainer request for AgentKit use-case docs, and the latest PR head commit was committed by Kevin Lin. (role: reviewer and recent PR committer; confidence: medium; commits: f09bc5a65652; files: docs/plugins/building-plugins.md, docs/plugins/hooks.md)

What the crustacean ranks mean

🦀 challenger crab: rare, exceptional readiness with strong proof, clean implementation, and convincing validation.
🦞 diamond lobster: very strong readiness with only minor maintainer review expected.
🐚 platinum hermit: good normal PR, likely mergeable with ordinary maintainer review.
🦐 gold shrimp: useful signal, but proof or patch confidence is still limited.
🦪 silver shellfish: thin signal; proof, validation, or implementation needs work.
🧂 unranked krab: not merge-ready because proof is missing/unusable or there are serious correctness or safety concerns.
🌊 off-meta tidepool: rating does not apply to this item.

Shiny media proof means a screenshot, video, or linked artifact directly shows the changed behavior. Runtime, network, CSP, and security claims still need visible diagnostics.

How this review workflow works

ClawSweeper keeps one durable marker-backed review comment per issue or PR.
Re-runs edit this comment so the latest verdict, findings, and automation markers stay together instead of adding duplicate bot comments.
A fresh review can be triggered by eligible @clawsweeper re-review comments, exact-item GitHub events, scheduled/background review runs, or manual workflow dispatch.
PR/issue authors and users with repository write access can comment @clawsweeper re-review or @clawsweeper re-run on an open PR or issue to request a fresh review only.
Maintainers can also comment @clawsweeper review to request a fresh review only.
Fresh-review commands do not start repair, autofix, rebase, CI repair, or automerge.
Maintainer-only repair and merge flows require explicit commands such as @clawsweeper autofix, @clawsweeper automerge, @clawsweeper fix ci, or @clawsweeper address review.
Maintainers can comment @clawsweeper explain to ask for more context, or @clawsweeper stop to stop active automation.

Guardiola31337 · 2026-05-17T20:56:45Z

@clawsweeper re-review

clawsweeper · 2026-05-17T20:58:13Z

🦞🧹
ClawSweeper re-review requested.

I asked ClawSweeper to review this item again.
Action: item re-review queued (workflow sweep.yml, event repository_dispatch).
Result: the existing ClawSweeper review comment will be edited in place when the review finishes.

Re-review progress:

State: Complete
Detail: The targeted re-review finished, the durable review comment was updated, and the synced verdict was routed.
Run: https://github.com/openclaw/clawsweeper/actions/runs/26002531898
Updated: 2026-05-17T21:06:44.592Z

Guardiola31337 · 2026-05-17T21:31:05Z

@clawsweeper re-review

I pushed e2b15b4606 to address the Web UI action rendering finding. The Web UI parser now preserves plugin approval actions and allowedDecisions, renders plugin command actions as command rows, and only renders decision buttons for decisions the Gateway can resolve.

Local proof after the patch:

$ node scripts/run-vitest.mjs ui/src/ui/controllers/exec-approval.test.ts ui/src/ui/views/exec-approval.test.ts ui/src/ui/app-gateway.node.test.ts src/gateway/server-methods/plugin-approval.test.ts src/infra/approval-view-model.test.ts src/plugin-sdk/approval-renderers.test.ts
Test Files  7 passed (7)
Tests  142 passed (142)

$ pnpm protocol:check
(exit 0)

$ pnpm plugin-sdk:api:check
OK docs/.generated/plugin-sdk-api-baseline.sha256

$ git diff --check
(no output)

clawsweeper · 2026-05-17T21:32:40Z

🦞🧹
ClawSweeper re-review requested.

I asked ClawSweeper to review this item again.
Action: item re-review queued (workflow sweep.yml, event repository_dispatch).
Result: the existing ClawSweeper review comment will be edited in place when the review finishes.

Re-review progress:

State: Complete
Detail: The targeted re-review finished, the durable review comment was updated, and the synced verdict was routed.
Run: https://github.com/openclaw/clawsweeper/actions/runs/26003309428
Updated: 2026-05-17T21:38:04.441Z

clawsweeper · 2026-05-20T15:10:13Z

ClawSweeper PR egg

🎁 Pass real behavior proof to wake the egg and unlock a hatchable treat.

Where did the egg go?

The egg game starts only after the PR passes the real-behavior proof check.
Before that, no creature or rarity is rolled. The treat waits for real proof.
This is still just collectible flavor: proof affects review readiness, not creature quality.

Guardiola31337 · 2026-05-21T14:44:44Z

@clawsweeper hatch

clawsweeper · 2026-05-21T14:46:43Z

🦞👀
ClawSweeper PR egg hatch requested.

I queued a comment sync for this PR. If the egg is hatchable, ClawSweeper will generate the image once and update the existing review comment.
Action: PR egg hatch queued (workflow sweep.yml, event repository_dispatch).
The ASCII egg stays as the fallback.

Guardiola31337 · 2026-05-24T14:23:23Z

@clawsweeper re-review

I pushed 22bc3fa246 to address the native approval callback finding. Slack and Telegram now render plugin command actions as visible command text, while decision actions use canonical OpenClaw /approve <id> <decision> callback payloads for native buttons, independent of plugin-provided command templates.

Proof:

node scripts/run-vitest.mjs extensions/telegram/src/approval-handler.runtime.test.ts extensions/slack/src/approval-handler.runtime.test.ts src/infra/approval-view-model.test.ts src/plugin-sdk/approval-renderers.test.ts
node scripts/run-oxlint.mjs --tsconfig config/tsconfig/oxlint.extensions.json extensions/telegram/src/approval-handler.runtime.ts extensions/telegram/src/approval-handler.runtime.test.ts extensions/slack/src/approval-handler.runtime.ts extensions/slack/src/approval-handler.runtime.test.ts
node scripts/run-tsgo.mjs -p test/tsconfig/tsconfig.core.test.json --incremental --tsBuildInfoFile /tmp/openclaw-82431-core-test-canonical-4.tsbuildinfo
node scripts/run-tsgo.mjs -p test/tsconfig/tsconfig.extensions.test.json --incremental --tsBuildInfoFile /tmp/openclaw-82431-extensions-test-canonical-4.tsbuildinfo
pnpm plugin-sdk:api:check
git diff --check

clawsweeper · 2026-05-24T14:23:26Z

🦞🧹
ClawSweeper re-review requested.

I asked ClawSweeper to review this item again.
Action: item re-review queued (workflow sweep.yml, event repository_dispatch).
Result: the existing ClawSweeper review comment will be edited in place when the review finishes.

Re-review progress:

State: Complete
Detail: The targeted re-review finished, the durable review comment was updated, and the synced verdict was routed.
Run: https://github.com/openclaw/clawsweeper/actions/runs/26363814619
Updated: 2026-05-24T14:34:14.381Z

kevinslin · 2026-05-27T02:32:50Z

this is a big change to our approval/plugin surface area. do you have some docs to go over the specific usecases in agentkit that you'd want to enable with this?

Guardiola31337 · 2026-05-27T15:38:29Z

Thank you so much for the review @kevinslin, and for landing the PR! 😊

Yep, I added the use-case context + diagrams to the PR descriptions, but they may be buried under the verification details.

Short version: this is for the external AgentKit ClawHub plugin, not bundled core code:

Flow: protected tool -> OpenClaw plugin approval -> AgentKit renders World verification -> proof succeeds -> plugin resolves only its own pending approval -> agent continues.

The remaining PRs are the generic host seams for that:

feat: add external plugin approval verification #82434: plugin-owned verified approval resolution
Expose chat.inject metadata for plugin approval updates #82471: durable/idempotent approval status cards in chat/history

Happy to pull the PR-body diagrams into a shorter standalone doc if that would make review easier.

openclaw-barnacle Bot added docs Improvements or additions to documentation app: web-ui App: web-ui gateway Gateway runtime agents Agent runtime and tooling size: L triage: needs-real-behavior-proof Candidate: external PR needs after-fix proof from a real setup. labels May 16, 2026

clawsweeper Bot added the mantis: telegram-visible-proof Mantis should capture Telegram visible proof. label May 16, 2026

Guardiola31337 mentioned this pull request May 16, 2026

feat: add external plugin approval verification #82434

Closed

Guardiola31337 force-pushed the agentkit/minimal-approval-actions branch from 5bf5b5a to 084316e Compare May 16, 2026 04:29

Guardiola31337 force-pushed the agentkit/minimal-approval-actions branch from 084316e to b41a29c Compare May 16, 2026 04:43

clawsweeper Bot added the proof: sufficient ClawSweeper judged the real behavior proof convincing. label May 16, 2026

openclaw-barnacle Bot removed the proof: sufficient ClawSweeper judged the real behavior proof convincing. label May 16, 2026

Guardiola31337 force-pushed the agentkit/minimal-approval-actions branch from 9aa6d85 to b1e22ba Compare May 16, 2026 05:08

clawsweeper Bot added the proof: sufficient ClawSweeper judged the real behavior proof convincing. label May 16, 2026

openclaw-barnacle Bot removed the proof: sufficient ClawSweeper judged the real behavior proof convincing. label May 16, 2026

clawsweeper Bot added the proof: sufficient ClawSweeper judged the real behavior proof convincing. label May 16, 2026

Guardiola31337 force-pushed the agentkit/minimal-approval-actions branch from 5477cff to 478115b Compare May 16, 2026 05:53

openclaw-barnacle Bot removed the proof: sufficient ClawSweeper judged the real behavior proof convincing. label May 16, 2026

clawsweeper Bot added the proof: sufficient ClawSweeper judged the real behavior proof convincing. label May 16, 2026

Guardiola31337 force-pushed the agentkit/minimal-approval-actions branch from 478115b to 2c35aad Compare May 16, 2026 06:09

openclaw-barnacle Bot removed the proof: sufficient ClawSweeper judged the real behavior proof convincing. label May 16, 2026

Guardiola31337 force-pushed the agentkit/minimal-approval-actions branch 2 times, most recently from cd3909a to de97b65 Compare May 16, 2026 06:17

openclaw-barnacle Bot added size: XL and removed proof: sufficient ClawSweeper judged the real behavior proof convincing. labels May 17, 2026

clawsweeper Bot added proof: sufficient ClawSweeper judged the real behavior proof convincing. impact:security Security boundary, credential, authz, sandbox, or sensitive-data risk. impact:session-state Session, memory, transcript, context, or agent state can drift or corrupt. labels May 17, 2026

Guardiola31337 force-pushed the agentkit/minimal-approval-actions branch from e42afcb to 4f315fc Compare May 17, 2026 20:53

openclaw-barnacle Bot removed the proof: sufficient ClawSweeper judged the real behavior proof convincing. label May 17, 2026

clawsweeper Bot added the proof: sufficient ClawSweeper judged the real behavior proof convincing. label May 17, 2026

openclaw-barnacle Bot removed the proof: sufficient ClawSweeper judged the real behavior proof convincing. label May 17, 2026

clawsweeper Bot added the proof: sufficient ClawSweeper judged the real behavior proof convincing. label May 17, 2026

Guardiola31337 force-pushed the agentkit/minimal-approval-actions branch from e2b15b4 to 1d71293 Compare May 18, 2026 12:09

openclaw-barnacle Bot removed the proof: sufficient ClawSweeper judged the real behavior proof convincing. label May 18, 2026

Guardiola31337 force-pushed the agentkit/minimal-approval-actions branch from 1d71293 to 67ca0a8 Compare May 18, 2026 12:12

clawsweeper Bot added the proof: sufficient ClawSweeper judged the real behavior proof convincing. label May 18, 2026

github-actions Bot mentioned this pull request May 27, 2026

🦞 OpenClaw 生态日报 2026-05-27 ivanweng2077/big_model_radar#97

Open

feat: expose plugin approval action metadata

f09bc5a

kevinslin mentioned this pull request May 27, 2026

revert: expose plugin approval action metadata #87419

Merged

This was referenced Jun 18, 2026

Expose plugin approval APIs needed by external HITL plugins #82336

Open

fix(types): unblock changed gate checks #87695

Merged

Uh oh!

Uh oh!

Conversation

Guardiola31337 commented May 16, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Review context

Real behavior proof

External plugin integration proof

Uh oh!

clawsweeper Bot commented May 16, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Guardiola31337 commented May 17, 2026

Uh oh!

clawsweeper Bot commented May 17, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Guardiola31337 commented May 17, 2026

Uh oh!

clawsweeper Bot commented May 17, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

clawsweeper Bot commented May 20, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Guardiola31337 commented May 21, 2026

Uh oh!

clawsweeper Bot commented May 21, 2026

Uh oh!

Guardiola31337 commented May 24, 2026

Uh oh!

clawsweeper Bot commented May 24, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

kevinslin commented May 27, 2026

Uh oh!

Guardiola31337 commented May 27, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Guardiola31337 commented May 16, 2026 •

edited

Loading

clawsweeper Bot commented May 16, 2026 •

edited

Loading

clawsweeper Bot commented May 17, 2026 •

edited

Loading

clawsweeper Bot commented May 17, 2026 •

edited

Loading

clawsweeper Bot commented May 20, 2026 •

edited

Loading

clawsweeper Bot commented May 24, 2026 •

edited

Loading