mirrors/hermes-agent
1
0
Fork 0
mirror of https://github.com/NousResearch/hermes-agent.git synced 2026-07-11 13:41:53 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 5

fix(caching): honor prompt_caching.enabled across model switch + fallback

Browse source 
@janrenz's PR #35862 added prompt_caching.enabled=false at init only. But
_anthropic_prompt_cache_policy re-derives _use_prompt_caching on every /model
switch (agent_runtime_helpers) and fallback-model swap (chat_completion_helpers),
which re-enabled markers and re-broke the strict proxy the toggle was meant to fix.

Move the kill switch into anthropic_prompt_cache_policy so it returns (False, False)
on every path. Drop the now-redundant init-time override (kept @janrenz's isinstance
hardening on the cache_ttl read). Add policy-level tests + docs for the toggle.

Follow-up to salvaged PR #35862.
This commit is contained in:
teknium1 2026-06-30 23:32:27 -07:00 • committed by Teknium
parent c1c1a12fe6
commit 36f9f50145
6 changed files with 109 additions and 6 deletions
Download patch file Download diff file Expand all files Collapse all files

1
scripts/release.py
View file

@ -45,6 +45,7 @@ ACP_REGISTRY_MANIFEST = REPO_ROOT / "acp_registry" / "agent.json"
# Auto-extracted from noreply emails + manual overrides
AUTHOR_MAP = {
"janrenz@Mac.fritz.box": "janrenz", # PR #35862 salvage (prompt_caching.enabled escape hatch for strict providers)
"syahidfrd@gmail.com": "syahidfrd", # PR #17059 salvage (tag unverified senders in Slack thread context to mitigate indirect prompt injection)
"5823452+sgabel@users.noreply.github.com": "sgabel", # PR #13139 salvage (redact secrets in user-facing approval prompts)
"130270192+CRWuTJ@users.noreply.github.com": "CRWuTJ", # PR #17082 salvage (cancel delayed Telegram deliveries on disconnect so buffered flushes don't dispatch into a torn-down session)