consciousness/poc-memory/agents
ProofOfConcept e12dea503b agent evaluate: sort agent actions by quality using Vec::sort_by with LLM
Yes, really. Rust's stdlib sort_by with an LLM pairwise comparator.
Each comparison is an API call asking "which action was better?"

Sample N actions per agent type, throw them all in a Vec, sort.
Where each agent's samples cluster = that agent's quality score.
Reports per-type average rank and quality ratio.

Supports both haiku (fast/cheap) and sonnet (quality) as comparator.

Usage: poc-memory agent evaluate --samples 5 --model haiku

Co-Authored-By: Kent Overstreet <kent.overstreet@linux.dev>
2026-03-14 19:24:07 -04:00
..
challenger.agent add {{siblings}} placeholder for graph neighborhood context 2026-03-12 18:08:58 -04:00
compare.agent agent evaluate: sort agent actions by quality using Vec::sort_by with LLM 2026-03-14 19:24:07 -04:00
connector.agent agents: port knowledge agents to .agent files with visit tracking 2026-03-10 17:04:44 -04:00
digest.agent digest: structural links, story-like prompt, agent file 2026-03-13 21:37:56 -04:00
distill.agent agents: add distill agent for core concept maintenance 2026-03-14 19:14:54 -04:00
evaluate.agent agents: add evaluate agent stub, fix distill query 2026-03-14 19:16:47 -04:00
extractor.agent agents: surface search hit counts to guide keep/demote decisions 2026-03-11 00:18:58 -04:00
health.agent agents: extract shared run_one_agent, standardize output formats 2026-03-10 17:33:12 -04:00
linker.agent agents: encourage hub creation and knowledge percolation 2026-03-14 17:21:07 -04:00
naming.agent naming agent: resolve node names before creation 2026-03-10 23:23:14 -04:00
observation.agent agents: port knowledge agents to .agent files with visit tracking 2026-03-10 17:04:44 -04:00
organize.agent agents: encourage hub creation and knowledge percolation 2026-03-14 17:21:07 -04:00
rename.agent rename agent: keys are concepts, update naming conventions 2026-03-14 18:52:35 -04:00
replay.agent agents: extract shared run_one_agent, standardize output formats 2026-03-10 17:33:12 -04:00
separator.agent agents: extract shared run_one_agent, standardize output formats 2026-03-10 17:33:12 -04:00
split.agent agents: placeholder-based prompt templates, port remaining 4 agents 2026-03-10 15:50:54 -04:00
transfer.agent add {{siblings}} placeholder for graph neighborhood context 2026-03-12 18:08:58 -04:00