7 Commits

Author	SHA1	Message	Date
uiyzzi	16d23d8cdc	feat(security): add sensitive data filtering for tool results sent to ... LLM Prevent LLM from seeing its own credentials (API keys, tokens, secrets) by filtering sensitive values from tool call results before sending to the model. Values are collected from .security.yml and replaced with [FILTERED] using an efficient strings.Replacer (O(n+m)). - Add FilterSensitiveData and FilterMinLength to ToolsConfig - Implement SensitiveDataReplacer() with sync.Once caching in SecurityConfig - Use reflection to collect all sensitive values (Model API keys, channel tokens, web tool API keys, skills tokens) - Apply filtering in agent loop at 4 tool result locations - Add comprehensive tests covering all token types	2026-03-23 20:55:41 +08:00
Cytown	5a8aab8143	Merge branch 'main' into version	2026-03-23 11:41:36 +08:00
Cytown	310f788f5f	rename security.yml to .security.yml	2026-03-23 11:20:42 +08:00
Cytown	284ced1f5c	Merge branch 'main' into version	2026-03-22 19:58:33 +08:00
Cytown	3dfe484f66	make yaml indent with 2	2026-03-22 11:07:22 +08:00
Cytown	7c854fe6d7	Merge branch 'main' into version	2026-03-22 02:53:55 +08:00
Cytown	e455eb5e67	refactor: seperate security.yml for store keys	2026-03-22 01:55:00 +08:00