Summary

• Problem: Telegram splits structured content (Markdown, heredoc, ### headings) earlier than the hard 4096-char limit. First fragments around ~3900 chars were not recognized as "near-limit" because the aggregation threshold was set at 4000 chars.
• Why it matters: Users pasting long structured messages in Telegram saw each fragment processed as a separate message with separate replies, instead of being aggregated into one.
• What changed: Lowered TELEGRAM_TEXT_FRAGMENT_START_THRESHOLD_CHARS from 4000 to 3600 in src/telegram/bot-handlers.ts.
• What did NOT change: Fragment timing, max gap, max parts, or total chars limits remain the same. Only the initial detection threshold is lowered.

Change Type (select all)

• Bug fix

Scope (select all touched areas)

• Integrations

Linked Issue/PR

• Closes [Bug]: Telegram inbound long-message aggregation fails when first fragment is below near-limit threshold (~4000 chars) #26237

User-visible / Behavior Changes

Telegram long-message aggregation now triggers reliably for structured content that Telegram splits at ~3800-4000 chars, instead of only triggering when the first fragment exceeds 4000 chars.

Security Impact (required)

• New permissions/capabilities? No
• Secrets/tokens handling changed? No
• New/changed network calls? No
• Command/tool execution surface changed? No
• Data access scope changed? No

Repro + Verification

Environment

• Integration/channel: Telegram

Steps

1. Send a long structured Markdown message (~4000+ chars) in Telegram
2. Telegram splits it into 2+ fragments
3. Observe that OpenClaw aggregates all fragments into one message

Expected

• Single aggregated message processed by the agent

Actual (before fix)

• Each fragment processed separately with separate replies

Evidence

• Failing test/log before + passing after
• New test "aggregates fragments when first part is ~3900 chars" added and passing

Human Verification (required)

• Verified scenarios: New test with 3900-char first fragment confirms aggregation works
• Edge cases checked: Existing tests with 4050-char fragments still pass (above new threshold)
• What you did not verify: Live Telegram environment testing

Compatibility / Migration

• Backward compatible? Yes
• Config/env changes? No
• Migration needed? No

Failure Recovery (if this breaks)

• How to disable/revert this change quickly: Revert the single constant back to 4000
• Files/config to restore: src/telegram/bot-handlers.ts
• Known bad symptoms: Short independent messages (~3600-4000 chars) being incorrectly buffered

Risks and Mitigations

• Risk: Independent short messages (3600-4000 chars) might be incorrectly buffered
  • Mitigation: Aggregation also requires sequential message IDs within the timing gap; standalone messages flush after the gap timer (1500ms default)

Greptile Summary

Lowered Telegram text fragment aggregation threshold from 4000 to 3600 chars to catch structured messages that Telegram splits earlier than the hard limit.

• Modified TELEGRAM_TEXT_FRAGMENT_START_THRESHOLD_CHARS constant in src/telegram/bot-handlers.ts:100
• Added test case covering 3900-char first fragment (previously below threshold)
• Existing test with 4050-char fragment still passes (above new threshold)

Confidence Score: 5/5

• Safe to merge - single constant change with comprehensive test coverage
• Simple threshold adjustment (4000→3600) with clear rationale, no logic changes, and test coverage for both old and new threshold boundaries. Aggregation safety preserved via timing gap and message ID checks.
• No files require special attention

_{Last reviewed commit: 0594daa}