# NousResearch/hermes-agent · Improve signal retention during context compression

## Open now

- 2 issues, 1 PR open on GitHub.
- #37010: Pre-Compression Memory Extraction with Quality Filtering and Deduplication (Working Implementation)
- #57875: \[Enhancement\]: Preserve high-signal decisions during repeated context compression
- PR #62777: feat(compression): preserve fidelity across repeated compactions
- _Broad area: open items may be separate problems._

These items explore enhancements to context compression pipelines, focusing on preserving high-quality or high-signal information across repeated compactions while filtering noise and deduplicating content. Approaches include pre-compression memory extraction, quality filtering, and LLM-based pre-filtering to reduce token cost.

## TL;DR

- 4 member(s): 2 open issue(s), 1 closed issue(s), 1 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evidence (`STRUCTURAL_PARTIAL`): some members are connected by structural evidence (shared changed files, or a closing/superseded reference); the rest are joined by an automated relevance judgement over their text.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 `0.29`: 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

- #37010: Pre-Compression Memory Extraction with Quality Filtering and Deduplication (Working Implementation)
- #57875: \[Enhancement\]: Preserve high-signal decisions during repeated context compression

## Closed / duplicate issues

- #16834: feat(hindsight): LLM-based retain pre-filter to reduce noise and token cost

## Prior & competing PRs

- **#62777: feat(compression): preserve fidelity across repeated compactions**. open
  - maintainer direction: category `uncategorized`; implication `needs_llm`
  - files: `agent/context_compressor.py`, `tests/agent/test_context_compressor_fidelity_ledger.py`, `tests/agent/test_context_compressor_signal_anchors.py`

## Fix status

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