The agent that grows with you
Python aiai-agentai-agentsanthropicchatgptclaudeclaude-codecodexhermeshermes-agentllmnous-researchopenai
plugin-catalog: re-pin agora to v2.0.13 (fix #11: motions failed with error/0 steps on Hermes v0.22.5)
Ethan Carter1ce6d3261d24
diff --git a/plugin-catalog/agora.yaml b/plugin-catalog/agora.yamlindex 10b4c6096dc..12c04d04aaa 100644--- a/plugin-catalog/agora.yaml+++ b/plugin-catalog/agora.yaml@@ -1,6 +1,6 @@name: agorarepo: https://github.com/yzy806806/agora-sha: 4e06ea3daa38aabf280483a91bbba64315927369+sha: 64e36e5cb51f8e2d1f06a510a637b93ef8065fabdescription: >-Self-driving multi-agent development team. Role-based workers (architect, developer,reviewer, tester, researcher, writer, devops) are each a real Hermes profile with its@@ -17,8 +17,9 @@ description: >---clone`, which copies config.yaml and .env into the new profile as owner-only files,never symlinks to the real credentials); spawns`hermes -p <profile> chat -Q -q` subprocesses to speak in- discussions and to run dispatched tasks, which execute as full agents with that profile's- toolset, including terminal and file writes inside the project workdir. Those subprocesses+ discussions (explicit toolset list: no kanban/clarify/cron; override via AGORA_SPEAKER_TOOLSETS)+ and to run dispatched tasks, which execute as full agents with that profile's toolset,+ including terminal and file writes inside the project workdir. Those subprocessesreceive `--yolo --accept-hooks` ONLY when the project opted in via`agora_start_project(allow_unattended=True)`; the default is off, so a freshly installedteam can read, discuss and plan but its flagged writes and shell commands fail closed.@@ -36,7 +37,7 @@ tier: communitycategory: automationrequires_hermes: '>=0.20'docs_url: https://github.com/yzy806806/agora#readme-version: "2.0.12"+version: "2.0.13"platforms: []readme: truecapabilities: