fix(compressor): hardening — max_tokens cap, multimodal safety, note fix, adaptive cooldown#9678
Closed
kshitijk4poor wants to merge 1 commit into
Closed
Conversation
…fix, adaptive cooldown
7 tasks
teknium1
pushed a commit
that referenced
this pull request
Apr 15, 2026
…rade, hardening Combined salvage of PRs #9661, #9663, #9674, #9677, #9678 by kshitijk4poor. - Smart tool output collapse: informative 1-line summaries replace generic placeholder - Dedup identical tool results via MD5 hash, truncate large tool_call arguments - Anti-thrashing: skip compression after 2 consecutive <10% savings passes - Structured action-log summary template with numbered actions and Active State - Hardening: max_tokens 1.3x cap, multimodal safety, note idempotency, adaptive cooldown Follow-up fixes applied during salvage: - web_extract: reads 'urls' (list) not 'url' (original PR bug) - Multimodal list content guards in dedup and prune passes - Kept 'Relevant Files' section in template (original PR removed it) Skipped PRs #9665 (user msg preservation — duplication risk) and #9675 (dead code).
teknium1
pushed a commit
that referenced
this pull request
Apr 15, 2026
…rade, hardening Combined salvage of PRs #9661, #9663, #9674, #9677, #9678 by kshitijk4poor. - Smart tool output collapse: informative 1-line summaries replace generic placeholder - Dedup identical tool results via MD5 hash, truncate large tool_call arguments - Anti-thrashing: skip compression after 2 consecutive <10% savings passes - Structured action-log summary template with numbered actions and Active State - Hardening: max_tokens 1.3x cap, multimodal safety, note idempotency, adaptive cooldown Follow-up fixes applied during salvage: - web_extract: reads 'urls' (list) not 'url' (original PR bug) - Multimodal list content guards in dedup and prune passes - Kept 'Relevant Files' section in template (original PR removed it) Skipped PRs #9665 (user msg preservation — duplication risk) and #9675 (dead code).
Contributor
|
Merged via #10088. Your changes were salvaged onto current main with your authorship preserved. Thanks for the contribution! 🎉 |
ulasbilgen
pushed a commit
to ulasbilgen/hermes-adhd-agent
that referenced
this pull request
May 1, 2026
…rade, hardening Combined salvage of PRs NousResearch#9661, NousResearch#9663, NousResearch#9674, NousResearch#9677, NousResearch#9678 by kshitijk4poor. - Smart tool output collapse: informative 1-line summaries replace generic placeholder - Dedup identical tool results via MD5 hash, truncate large tool_call arguments - Anti-thrashing: skip compression after 2 consecutive <10% savings passes - Structured action-log summary template with numbered actions and Active State - Hardening: max_tokens 1.3x cap, multimodal safety, note idempotency, adaptive cooldown Follow-up fixes applied during salvage: - web_extract: reads 'urls' (list) not 'url' (original PR bug) - Multimodal list content guards in dedup and prune passes - Kept 'Relevant Files' section in template (original PR removed it) Skipped PRs NousResearch#9665 (user msg preservation — duplication risk) and NousResearch#9675 (dead code).
aj-nt
pushed a commit
to aj-nt/hermes-agent
that referenced
this pull request
May 1, 2026
…rade, hardening Combined salvage of PRs NousResearch#9661, NousResearch#9663, NousResearch#9674, NousResearch#9677, NousResearch#9678 by kshitijk4poor. - Smart tool output collapse: informative 1-line summaries replace generic placeholder - Dedup identical tool results via MD5 hash, truncate large tool_call arguments - Anti-thrashing: skip compression after 2 consecutive <10% savings passes - Structured action-log summary template with numbered actions and Active State - Hardening: max_tokens 1.3x cap, multimodal safety, note idempotency, adaptive cooldown Follow-up fixes applied during salvage: - web_extract: reads 'urls' (list) not 'url' (original PR bug) - Multimodal list content guards in dedup and prune passes - Kept 'Relevant Files' section in template (original PR removed it) Skipped PRs NousResearch#9665 (user msg preservation — duplication risk) and NousResearch#9675 (dead code).
02356abc
pushed a commit
to 02356abc/hermes-agent
that referenced
this pull request
May 14, 2026
…rade, hardening Combined salvage of PRs NousResearch#9661, NousResearch#9663, NousResearch#9674, NousResearch#9677, NousResearch#9678 by kshitijk4poor. - Smart tool output collapse: informative 1-line summaries replace generic placeholder - Dedup identical tool results via MD5 hash, truncate large tool_call arguments - Anti-thrashing: skip compression after 2 consecutive <10% savings passes - Structured action-log summary template with numbered actions and Active State - Hardening: max_tokens 1.3x cap, multimodal safety, note idempotency, adaptive cooldown Follow-up fixes applied during salvage: - web_extract: reads 'urls' (list) not 'url' (original PR bug) - Multimodal list content guards in dedup and prune passes - Kept 'Relevant Files' section in template (original PR removed it) Skipped PRs NousResearch#9665 (user msg preservation — duplication risk) and NousResearch#9675 (dead code).
gweeteve
pushed a commit
to gweeteve/hermes-agent
that referenced
this pull request
Jun 2, 2026
…rade, hardening Combined salvage of PRs NousResearch#9661, NousResearch#9663, NousResearch#9674, NousResearch#9677, NousResearch#9678 by kshitijk4poor. - Smart tool output collapse: informative 1-line summaries replace generic placeholder - Dedup identical tool results via MD5 hash, truncate large tool_call arguments - Anti-thrashing: skip compression after 2 consecutive <10% savings passes - Structured action-log summary template with numbered actions and Active State - Hardening: max_tokens 1.3x cap, multimodal safety, note idempotency, adaptive cooldown Follow-up fixes applied during salvage: - web_extract: reads 'urls' (list) not 'url' (original PR bug) - Multimodal list content guards in dedup and prune passes - Kept 'Relevant Files' section in template (original PR removed it) Skipped PRs NousResearch#9665 (user msg preservation — duplication risk) and NousResearch#9675 (dead code).
Egavasyug
pushed a commit
to Egavasyug/hermes-agent
that referenced
this pull request
Jun 10, 2026
…rade, hardening Combined salvage of PRs NousResearch#9661, NousResearch#9663, NousResearch#9674, NousResearch#9677, NousResearch#9678 by kshitijk4poor. - Smart tool output collapse: informative 1-line summaries replace generic placeholder - Dedup identical tool results via MD5 hash, truncate large tool_call arguments - Anti-thrashing: skip compression after 2 consecutive <10% savings passes - Structured action-log summary template with numbered actions and Active State - Hardening: max_tokens 1.3x cap, multimodal safety, note idempotency, adaptive cooldown Follow-up fixes applied during salvage: - web_extract: reads 'urls' (list) not 'url' (original PR bug) - Multimodal list content guards in dedup and prune passes - Kept 'Relevant Files' section in template (original PR removed it) Skipped PRs NousResearch#9665 (user msg preservation — duplication risk) and NousResearch#9675 (dead code).
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
2 participants
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.
Summary
Four small hardening fixes in
context_compressor.py, each addressing a concrete edge case.Fixes
1. Summary max_tokens overshoot (line 453)
Before:
max_tokens = summary_budget * 2— the summary model can generate 2x what was requested, wasting time and money on a 12K-token ceiling that becomes 24K.After:
max_tokens = int(summary_budget * 1.3)— 30% headroom for slight overshoots without doubling the cost.2. Multimodal content safety (line 214)
Before:
len(msg.get("content") or "")— crashes withTypeErrorif content is a list (multimodal messages with image blocks).After: Detects list content and sums text part lengths. Falls back to
len()for strings.3. Compression note lost after re-compression (line 747)
Before:
if self.compression_count == 0— note only added on first compression. If system prompt is rebuilt between compressions (which_compress_contextdoes), the note from #1 is lost and never re-added.After: Check if note already exists in the system prompt content — idempotent, survives rebuilds.
4. Adaptive failure cooldown (lines 468-483)
Before: Both
RuntimeError(no provider — permanent) and genericException(timeout/network — transient) get the same 600s cooldown. A 1-second network blip = 10 minutes of degraded compression.After:
RuntimeError→ 600s (permanent config issue). Generic exceptions → 60s (transient, likely recovers quickly).Test plan
Part of #9666.