Tool-loop guardrail improvements for content repetition

NousResearch/hermes-agent · cluster 60084 · agent

Data as of

2 open items. Fix attempted, not merged.

The short answer

2open issues
0open PRs
7items grouped
Known?
Yes. 7 related items grouped here, 0 issues already closed.
Fix event
No fix has been merged to main. 5 PRs closed without merging; see the attempts below.
Tracker
The maintainers' tracker marks this unresolved. That is separate from the fix event.
Still open
2 issues and 0 PRs.

These issues and PRs address gaps in tool-call loop detection, specifically cases where arguments vary but tool results remain the same or repeat. Work spans bug reports, warning-first approaches, and fixes for content-based repetition detection.

Open issues2

  • Issue#60084[Bug]: Tool-loop guardrails miss loops where args vary but the result never changes (incl. repeated multimodal results)
  • Issue#67889tool_loop_guardrails cannot detect thrashing on the terminal tool — three compounding gaps

Open PRs0

No open pull requests.

5 PRs closed without merging

From the close comments: duplicate 2 · stale or abandoned 2 · done elsewhere 1

  • #60087 feat(guardrails): detect repeated tool results independent of args/failure · done elsewhere → 37c27b6, #60084, #78551
  • #60148 fix(tool-guardrails): detect content repetition across varying argument sets (#60084) · stale or abandoned
  • #60219 fix(tool-guardrails): detect content-repetition loops across varying args (#60084) · stale or abandoned
  • #60268 fix(guardrails): detect content-based repetition loops even when args vary (#60084) · duplicate → #60087, #60084, #60219, #60094, #60148
  • #60285 fix(agent): detect same-result tool loops even when args vary · duplicate → #60087
How this cluster was grouped
  • Link evidence (partly structural links): some members are connected by structural evidence (shared changed files, or a closing/superseded reference); the rest are joined by an automated relevance judgement over their text.
  • Members outside that structural core may belong to a different underlying problem.
  • Pairwise link density 0.26: the share of item pairs that carry a link, counting a structural link in full and a relevance-only link at 0.36. 1.00 means every pair is linked.

Pick up a Lab task

Check whether a fix covers the open issues here, with your own agent. Bring back evidence for maintainers.

Open the Lab

At a glance

Subsystem
agent
Closed issues
0
Attempts
5 PRs closed without merging

For your agent

The same facts, typed.