fix(compressor): batch salvage \u2014 orphan tool_calls, cooldown abort flags, system-head summary role (#51225 #52056 #52167)

[hashbender]

Summary

User-requested batch salvage of three complementary P1 bug fixes to agent/context_compressor.py, rebased onto current main. All three targeted bugs are confirmed still present on main; the three fixes touch non-overlapping regions and compose cleanly.

Note: the original PR branches predated a since-landed cooldown-helper refactor (_record_/_clear_compression_failure_cooldown), so their raw diffs would have reverted that refactor. This salvage re-applies only the genuine fix intent on top of current main's helper-based code — the refactor is left intact (verified).

Included PRs (authorship preserved per commit)

• fix(compressor): strip orphan tool_calls instead of inserting stubs (#51218) NousResearch/hermes-agent#51225 (@liuhao1024) — fix(compressor): strip orphan tool_calls instead of inserting stubs (#51218)
_sanitize_tool_pairs inserted stub role="tool" results for orphaned tool_calls; the pre-API repair_message_sequence() keys on tc.get("id") while the sanitizer keys on call_id||id, so on the Codex Responses API format (id != call_id) the stubs were silently dropped and the orphans re-exposed. Now strips the orphaned tool_calls at the source (preserving text content, adding (tool call removed) only for an otherwise-empty assistant turn).

• fix(compression): preserve network/auth abort flags across cooldown re-entry (#29559) NousResearch/hermes-agent#52056 (@srojk34) — fix(compression): preserve network/auth abort flags across cooldown re-entry (#29559)
compress() eagerly reset _last_summary_auth_failure / _last_summary_network_failure at the top of every call. On a second compress() during the failure cooldown, _generate_summary() returns None from the cooldown early-return without re-asserting the flags, so the abort guard saw False and fell through to the destructive static-fallback that drops the middle window ([Bug] Context compression silently drops summary on connection error, causing total context loss for long tasks NousResearch/hermes-agent#29559/bug(compression): failed summaries should not discard conversation context NousResearch/hermes-agent#25585). The eager reset is removed; a successful summary already clears both flags, so persistence across calls is safe.

• fix(compressor): pin summary role to user when only system prompt is protected (#52160) NousResearch/hermes-agent#52167 (@liuhao1024) — fix(compressor): pin summary role to user when only system prompt is protected (#52160)
  After the first compaction protect_first_n decays, so a later compaction can have only the system prompt as the protected head. Anthropic/Bedrock send the system prompt as a separate parameter, making the summary the first messages[] entry — and Anthropic rejects a first message that isn't role=user (HTTP 400). Pins the summary to user for a system-only head and blocks the collision-flip from reverting it.

Follow-up commit (mine)

• test(compressor): tidy blank-line spacing + assert placeholder never overwrites text — review-driven test cleanup only, no production change.

Verification

• pytest tests/agent/test_context_compressor.py → 141 passed (includes 8 new tests across TestSanitizerStripsOrphanedToolCalls, TestCooldownReentryAbort, TestDoubleCompactionSummaryRole).
• Mutation-checked: reverting each of the three source fixes fails its corresponding tests — the tests are genuine behavior contracts, not change-detectors.
• Stale-revert check: the cooldown-helper refactor is intact (no raw _summary_failure_cooldown_until assignments reintroduced in compress()/_generate_summary).

Closes NousResearch#51225. Closes NousResearch#52056. Closes NousResearch#52167.

Co-authored-by: liuhao1024 sunsky.lau@gmail.com
Co-authored-by: srojk34 286497132+srojk34@users.noreply.github.com

---

Mirror-of: NousResearch#56197
NousResearch#56197

[tenki-reviewer]

Review Complete
No issues found!

Risk: 🟠 High (78/100) — no findings · 564 LOC across 2 files

---

The compression summary is silently dropped when _force_user_leading=True coincides with the Causal Coupling guard pushing compress_end to the end of the message list, causing data loss — the entire conversation after the system prompt is lost.

Files Reviewed (2 files)

agent/context_compressor.py
tests/agent/test_context_compressor.py