ShengtongZhu 2e2eed339a refactor(guardian): replace async instruction extraction with full system prompt caching ... Remove the LLM-based standingInstructions and availableSkills extraction pipeline. Instead, cache the main agent's full system prompt on the first llm_input and pass it as-is to the guardian as "Agent context". This eliminates two async LLM calls per session, simplifies the codebase (~340 lines removed), and gives the guardian MORE context (the complete system prompt including tool definitions, memory, and skills) rather than a lossy LLM-extracted summary. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>		2026-03-15 12:33:28 +08:00
..
directory.ts	fix(ci): repair helper typing regressions	2026-03-14 03:22:53 +00:00
plugin-api.ts	fix(ci): repair helper typing regressions	2026-03-14 03:22:53 +00:00
plugin-runtime-mock.ts	refactor(guardian): replace async instruction extraction with full system prompt caching	2026-03-15 12:33:28 +08:00
runtime-env.ts	Plugins/test-utils: migrate to scoped plugin-sdk imports	2026-03-04 02:35:13 -05:00
send-config.ts	test: share send cfg threading helpers	2026-03-14 02:40:27 +00:00
start-account-context.ts	Plugins/test-utils: migrate to scoped plugin-sdk imports	2026-03-04 02:35:13 -05:00
start-account-lifecycle.ts	test: share startup account lifecycle helpers	2026-03-14 01:41:17 +00:00
status-issues.ts	test: share status issue assertion helpers	2026-03-14 01:41:17 +00:00