36 lines
2.1 KiB
YAML
36 lines
2.1 KiB
YAML
name: hermes-switchyard
|
|
repo: https://github.com/bgrablin/hermes-switchyard
|
|
sha: b3cf54676aa0244c55b448fe84a2eb93061d91a8
|
|
subdir: ""
|
|
description: "Jev-powered Hermes plugin for skill/model routing, typed assess, DOM/Cua computer use, adaptive reasoning effort with a per-turn receipt line, and FTS rerank. Disclosure: the default-on adaptive llm_request sends up to 1,200 characters of the current user message to TypeSafe/OpenRouter after Hermes redact_for_egress (no history or tool output); greetings and thanks are decided locally with no call; without the Hermes redactor it sends metadata only (disable via adaptive_reasoning_effort false). Separate default-on hosted skill routing sends up to 4,000 characters of each user turn after Hermes redact_for_egress, plus skill names, to the same provider; without the redactor it stays local (opt down to local_only). Model routing remains advisory. Disclosure — Adaptive effort now sends up to 1,200 redacted characters of the current user message per turn by default (v0.5.3 sent metadata only); jev_session_search_rerank and jev_computer_use DOM runs send query/card text and public-page text (≤4,000 chars) to the same provider using the plugin's own masking rather than Hermes redact_for_egress; both are model-invoked tools gated on the default-true public_or_sanitized_data_ack."
|
|
maintainer: bgrablin
|
|
tier: community
|
|
category: tools
|
|
requires_hermes: ">=0.21.3"
|
|
docs_url: https://github.com/bgrablin/hermes-switchyard/blob/b3cf54676aa0244c55b448fe84a2eb93061d91a8/docs/SETUP.md
|
|
version: "0.5.6"
|
|
image: https://raw.githubusercontent.com/bgrablin/hermes-switchyard/b3cf54676aa0244c55b448fe84a2eb93061d91a8/docs/assets/hermes-switchyard-overview.png
|
|
readme: true
|
|
platforms:
|
|
- linux
|
|
- macos
|
|
- windows
|
|
capabilities:
|
|
provides_tools:
|
|
- jev_assess
|
|
- jev_computer_use
|
|
- jev_skill_select
|
|
- jev_skill_select_many
|
|
- jev_model_route
|
|
- jev_model_route_approved
|
|
- jev_session_search_rerank
|
|
provides_hooks:
|
|
- pre_llm_call
|
|
- post_llm_call
|
|
- post_tool_call
|
|
- transform_llm_output
|
|
- post_api_request
|
|
provides_middleware:
|
|
- llm_request
|
|
requires_env: []
|