hermes-agent/agent
teyrebaz33 119bad65fc feat: prune old tool outputs before context compaction
Port the useful part of PR #588 onto current main without regressing
summary role alternation or the centralized call_llm-based summary path.

This adds a prune-first compression pass that:
- protects recent tool outputs with adaptive thresholds
- never prunes key tool outputs like read_file/memory/clarify
- skips the LLM summary call entirely when pruning alone is enough
- keeps head/tail protected windows untouched

Tests cover prune-only compaction, protected tools, and tail protection.

Co-authored-by: teyrebaz33 <hakanerten02@hotmail.com>
2026-03-13 21:21:28 -07:00
..
__init__.py Refactor Terminal and AIAgent cleanup 2026-02-21 22:31:43 -08:00
anthropic_adapter.py Merge pull request #1149 from NousResearch/hermes/hermes-d28bf447 2026-03-13 03:09:43 -07:00
auxiliary_client.py merge: resolve conflicts with main (URL update to hermes-agent.nousresearch.com) 2026-03-12 17:49:26 -07:00
context_compressor.py feat: prune old tool outputs before context compaction 2026-03-13 21:21:28 -07:00
display.py Merge PR #736: feat(honcho): async writes, memory modes, session title integration, setup CLI 2026-03-12 19:05:11 -07:00
insights.py chore: add GLM/Kimi/MiniMax models to insights pricing (zero cost) 2026-03-06 19:12:14 -08:00
model_metadata.py fix(anthropic): live model fetching + adaptive thinking for 4.5+ models 2026-03-12 17:04:31 -07:00
prompt_builder.py feat: secure skill env setup on load (core #688) 2026-03-13 03:14:04 -07:00
prompt_caching.py Refactor Terminal and AIAgent cleanup 2026-02-21 22:31:43 -08:00
redact.py feat: secure skill env setup on load (core #688) 2026-03-13 03:14:04 -07:00
skill_commands.py feat: secure skill env setup on load (core #688) 2026-03-13 03:14:04 -07:00
trajectory.py Refactor Terminal and AIAgent cleanup 2026-02-21 22:31:43 -08:00