# NousResearch/hermes-agent · Parsing llama.cpp context-size errors for context-limit recovery

## Open now

- 1 issue, 2 PRs open on GitHub.
- #89502: Context-overflow recovery can't parse llama.cpp's \`exceed_context_size_error\`; sessions die with "Cannot compress further" when the resolved window overestimates the served \`-c\`
- PR #89504: fix(model_metadata): parse llama.cpp exceed_context_size_error for context-limit recovery
- PR #136396: llama.cpp overflow errors now teach Hermes the server's real context window (salvage #89504)

The three members share one theme: Hermes's context-overflow recovery reportedly cannot parse llama.cpp's \`exceed_context_size_error\`, and two members appear to propose changes to parse it and learn the server's actual context window. One issue describes sessions ending with "Cannot compress further" when the computed window exceeds the server's \`-c\` setting.

## TL;DR

- 3 member(s): 1 open issue(s), 0 closed issue(s), 2 open PR(s), 0 merged PR(s), 0 PR(s) closed without merging.
- No fix has been merged to main; one or more PRs were attempted (see below).
- **Recommended:** review the prior/competing PR(s) below. A fix was attempted but not merged; your case may extend or revive that work.

## Maintainer resolution verdict

- Internal-tracker verdict over this cluster's issue members: `unresolved`. This is distinct from the fix status above. Weigh it yourself.

## How this cluster was grouped

- Link evidence (`STRUCTURAL_SPANNING`): every member is connected by structural evidence: shared changed files, or a closing/superseded reference.
- Pairwise link density `0.67`: the share of member 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.

## Open related issues

- #89502: Context-overflow recovery can't parse llama.cpp's \`exceed_context_size_error\`; sessions die with "Cannot compress further" when the resolved window overestimates the served \`-c\`

## Prior & competing PRs

- **#89504: fix(model_metadata): parse llama.cpp exceed_context_size_error for context-limit recovery**. open
  - maintainer direction: category `duplicate_of`; implication `duplicate`; refs `#136396`, `#13421`
  - files: `agent/model_metadata.py`, `tests/agent/test_model_metadata.py`
- **#136396: llama.cpp overflow errors now teach Hermes the server's real context window (salvage #89504)**. open
  - maintainer direction: category `uncategorized`; implication `needs_llm`; refs `cc96dbfd7d62154527ddb2b61be8897b2503e73b`, `#13421`
  - files: `agent/model_metadata.py`, `tests/agent/test_model_metadata_llamacpp.py`

## Fix status

- No merged fix on record.
- Subsystem: `agent`