Thanks for the context here. I swept through the related work, and this is now duplicate or superseded.

Close as superseded: merged current-main work already replaced the raw-output and core thread-binding pieces, and the remaining Telegram forum-topic spawn behavior is tracked by the canonical open issue rather than this stale, broader branch.

Root-cause cluster
Relationship: superseded
Canonical: #42986
Summary: This PR overlaps merged current-main raw-output and core thread-binding fixes; the remaining Telegram forum-topic subagent spawn capability is tracked by the canonical issue and an active candidate PR.

Members:

• canonical: [Feature Request] Telegram Channel: Support Thread Binding for Subagent Sessions #42986 - Open issue tracks the remaining Telegram sessions_spawn(thread:true, mode:session) forum-topic binding capability.
• superseded: Fix Telegram forum-topic subagent completion delivery #85089 - Merged PR replaced the raw tool/toolResult output suppression portion of this branch.
• superseded: Refactor subagent thread binding into core #88416 - Merged PR moved native thread:true subagent binding into core, superseding this branch's hook/delivery-origin shape.
• fixed_by_candidate: feat(telegram): bind spawned subagents to forum topics #82023 - Open PR targets the remaining canonical Telegram forum-topic binding capability with live proof, but it is currently conflicting and not merged.
• partial_overlap: fix(agents): suppress raw subagent tool output #80110 - Open PR overlaps raw tool-output suppression while current main already contains broader completion-selector behavior.

Proposal only: this assessment does not dispatch repair, suppress jobs, mutate sibling items, close, or merge anything.

Canonical path: The remaining Telegram sessions_spawn(thread:true, mode:session) forum-topic capability is tracked by #42986; #82023 is an active but conflicting candidate for that issue.

So I’m closing this here and keeping the remaining discussion on #42986 and #82023.

Mantis Telegram Desktop Proof

Summary: Mantis did not generate before/after GIFs because PR 83274 changes Telegram-origin subagent thread-binding/tool availability and unsafe subagent announce filtering, but the native Telegram proof harness mock model cannot emit the required sessions_spawn(thread=true) tool call; a Telegram Desktop recording would only show a generic text reply or /status, not the PR behavior.

• Scenario: telegram-desktop-proof
• Trigger: pull_request_target
• Run: https://github.com/openclaw/openclaw/actions/runs/26003579850
• Artifact: https://github.com/openclaw/openclaw/actions/runs/26003579850/artifacts/7046443754
• Baseline: skipped at c93d6d8daa37fadbbd98611f08c7489b3dc14097, expected no visible Telegram Desktop delta
• Candidate: skipped at a860ec6a6481764cf7e067b2396d080996f4727c, expected no visible Telegram Desktop delta
• Overall: true

Raw QA files: https://artifacts.openclaw.ai/mantis/telegram-desktop/pr-83274/run-26003579850-1/index.json

@clawsweeper re-review

Addressed the review findings in 801308c:

• Removed the hardcoded Telegram special-case from core thread-binding spawn policy; Telegram now advertises child placement through its bundled thread-binding-api.ts public surface.
• Preserved indirect thread-bound session routing by resolving a bound child conversation when the hook reports ready without a direct deliveryOrigin, and only using direct delivery when that route is actually routable.
• Applied unsafe-output filtering to the fallback latest-assistant reply path, with regression coverage.

Verification run locally:

• OPENCLAW_VITEST_MAX_WORKERS=1 node scripts/run-vitest.mjs run --config test/vitest/vitest.agents.config.ts src/agents/subagent-announce-output.test.ts src/agents/subagent-announce.timeout.test.ts src/agents/subagent-spawn.thread-binding.test.ts src/agents/sessions-spawn-hooks.test.ts
• OPENCLAW_VITEST_MAX_WORKERS=1 node scripts/run-vitest.mjs run --config test/vitest/vitest.channels.config.ts src/channels/thread-bindings-policy.test.ts src/channels/plugins/thread-binding-api.test.ts
• corepack pnpm format:check ...
• git diff --check

🦞🧹
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/26014236153
• Updated: 2026-05-18T05:00:16.564Z

@clawsweeper re-review

Pushed follow-up commit ff9a55cc8a for the latest blockers:

• Kept Telegram placement consistent for loaded plugins and the bundled public artifact: extensions/telegram/src/channel.ts now advertises defaultTopLevelPlacement: "child", with a regression test comparing it to extensions/telegram/thread-binding-api.ts.
• Replaced the invalid binding status comparison with the real active-status check (binding.status === "active").
• Fixed the hook mock type regression by explicitly typing the test hook result so deliveryOrigin remains optional.

Local verification after the fix:

OPENCLAW_TSGO_MAX_WORKERS=1 corepack pnpm tsgo:core:test
OPENCLAW_TSGO_MAX_WORKERS=1 corepack pnpm tsgo:extensions:test
OPENCLAW_VITEST_MAX_WORKERS=1 node scripts/run-vitest.mjs run --config test/vitest/vitest.agents.config.ts src/agents/subagent-spawn.thread-binding.test.ts src/agents/sessions-spawn-hooks.test.ts
OPENCLAW_VITEST_MAX_WORKERS=1 node scripts/run-vitest.mjs run --config test/vitest/vitest.extension-telegram.config.ts extensions/telegram/src/thread-bindings.test.ts
OPENCLAW_VITEST_MAX_WORKERS=1 node scripts/run-vitest.mjs run --config test/vitest/vitest.channels.config.ts src/channels/thread-bindings-policy.test.ts src/channels/plugins/thread-binding-api.test.ts
OPENCLAW_TSGO_MAX_WORKERS=1 corepack pnpm tsgo:core
OPENCLAW_TSGO_MAX_WORKERS=1 corepack pnpm tsgo:extensions
corepack pnpm format:check src/agents/subagent-spawn.ts src/agents/sessions-spawn-hooks.test.ts src/agents/subagent-spawn.thread-binding.test.ts extensions/telegram/src/channel.ts extensions/telegram/thread-binding-api.ts extensions/telegram/src/thread-bindings.test.ts
git diff --check

All passed locally.

Real behavior proof is still pending: I do not have a safe redacted live Telegram runtime transcript from this checkout to attach yet. The latest commit addresses the concrete source/CI blockers from the review comment.

🦞🧹
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/26032280800
• Updated: 2026-05-18T12:13:18.189Z

ClawSweeper PR egg

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

@Svetznaniy33 thanks for the PR. ClawSweeper is still waiting on real behavior proof before this can move forward.

Useful proof can be a screenshot, short video, terminal output, copied live output, linked artifact, or redacted logs that show the changed behavior after the fix. Please redact private tokens, phone numbers, private endpoints, customer data, and anything else sensitive.

Once proof is added to the PR body or a comment, ClawSweeper or a maintainer can re-check it.

@Svetznaniy33 thanks for the PR. ClawSweeper is still waiting on real behavior proof before this can move forward.

Useful proof can be a screenshot, short video, terminal output, copied live output, linked artifact, or redacted logs that show the changed behavior after the fix. Please redact private tokens, phone numbers, private endpoints, customer data, and anything else sensitive.

Once proof is added to the PR body or a comment, ClawSweeper or a maintainer can re-check it.