feat(evolution): add self-evolution subsystem

Add a self-evolution subsystem that reviews idle conversations in an
isolated agent and durably learns from them — patching/creating skills,
finishing unfinished tasks, and backfilling missed memory.

- Trigger: background idle scan, fires when a session is idle >= N min AND
  (>= N turns OR context usage > 80%). In-memory cursor reviews only new
  messages so a session never re-learns old content.
- Isolated review agent: same model, restricted toolset, hard write-guard
  confining edits to the workspace (built-in skills are protected).
- Safety: file-level backup before edits + evolution_undo tool; notify the
  user ONLY when a workspace file actually changed (no-nag rule); capped
  concurrency.
- Records to memory/evolution/<date>.md, surfaced in the memory UI's
  renamed "Self-Evolution" tab (merged with dream diaries).
- Hide internal [SCHEDULED]/[EVOLUTION]/backup_id markers from chat history
  display (also fixes scheduler marker leakage) while keeping them in stored
  content for undo.
- Flat config: self_evolution_enabled (default off until release),
  self_evolution_idle_minutes (15), self_evolution_min_turns (6).
- Tests: tests/test_evolution.py (stub + real model modes, 7 scenarios).
This commit is contained in:
zhayujie
2026-06-07 18:55:33 +08:00
parent 0e4da1d1c5
commit ba777ed706
19 changed files with 1856 additions and 20 deletions

View File

@@ -251,6 +251,10 @@ available_setting = {
"enable_thinking": False, # Enable deep-thinking mode for thinking-capable models
"reasoning_effort": "high", # Reasoning depth under thinking mode: "high" or "max"
"knowledge": True, # whether to enable the knowledge base feature
# Self-evolution: review idle conversations to learn memory/skills. Flat keys.
"self_evolution_enabled": False, # master switch (off until release)
"self_evolution_idle_minutes": 15, # idle time before a session is reviewed
"self_evolution_min_turns": 6, # min user turns (or context pressure) to trigger
"skill": {}, # Per-skill runtime config; nested keys flatten to SKILL_<NAME>_<KEY> env vars at startup
"mcp_servers": [], # MCP server list; each entry supports type "stdio" (local process) or "sse" (remote URL)
}