escapeboy/agent-fleet-o

What's changed

• chore: release v1.17.0 — GPU templates, MCP marketplace, integrations (8f4839e)
• feat: add ToolTemplate REST API (index, show, deploy) (30565d4)
• feat: 1Password integration — MCP tool + native driver (95b5794)
• fix: remove hardcoded stats from marketplace seeder (d995052)
• feat: seed marketplace with 34 official listings (22 skills + 12 agents) (45937cb)
• feat: GPU templates, MCP marketplace, Apify, screenpipe, quick agent, integrations UX (78a3801)
• feat: AgentScope-inspired patterns — 5 features + test fixes (946ab4f)
• feat: a2ui MCP tools, migration, security audit, tests (88954ee)
• feat: GSD-2 inspired AI Gateway intelligence + full UI (e989be6)
• fix: KG search havingRaw bug + add chat memory extraction listener (9e1f86b)
• fix: add A2uiRenderer class (missed in a2ui surface commit) (136fb3c)
• fix: update tpetry/laravel-postgresql-enhanced 3.5→3.6 (tsvector compat fix) (1253c96)
• feat: add view/edit modals to Knowledge Graph browser page (b63027b)
• fix: add a2ui surface blade component (fixes view:cache on production) (7d2c902)
• fix: add a2ui surface blade component (fixes view:cache on production) (10c7c14)
• feat: assistant tools, MJML rendering, memory scoping, KG quality, artifact fixes (45d5a3d)
• fix(artifacts): resolve file references + add missing content types (15f0940)
• fix: remove 'team-knowledge' fallback — pass null for agentless experiments (397ace6)
• feat: assistant email theme tools, MJML CLI+HTTP rendering (977ec88)
• feat: assistant streaming, KG data quality, memory tag scoping (bf28b6e)
• fix(outbound): remove simulated sends, fix email target resolution, add agent/trigger/signal improvements (a11c6e4)
• feat(chatbot): add SSE streaming endpoint for chat messages (466374f)
• fix(chatbot): extract 'text' output key + fallback for all chatbot types (d9edd9e)
• fix(chatbot): lower RAG similarity threshold from 0.72 to 0.5 (b04ad96)
• fix(chatbot): add 'text' key to delivery output extraction (ea4c191)
• fix(chatbot): skill execution error messages + playbook step retry (6a9a13d)
• fix(chatbot): skill execution error messages + playbook step retry (5a4c077)
• feat(chatbot): add delivery listener for workflow results (370fd3a)
• fix(workflow): handle string input_mapping in resolveInput (0723ad4)
• fix(chatbot): skip scoring stage for chatbot workflow experiments (aa24389)
• fix(chatbot): use team owner_id instead of team_id as actorUserId (e8f6c95)
• fix(bridge): permanent stability improvements for unstable networks (1fba583)
• fix(security): IDOR guard on knowledge_base_id, authorization on saveAiFeatures, stage_model_tiers validation (9b016c2)
• feat: pre-existing config gaps — per-team overrides, GlobalSettings, agent fields (08c8ac6)
• feat: cloud configurability for Hermes-inspired features — per-team settings, UI, API, MCP (a5bb80d)
• fix: use experiment.thesis instead of non-existent goal column (9268982)

Upgrade

git fetch --tags origin
git checkout tags/v1.17.0
composer install --no-dev --optimize-autoloader
php artisan migrate --force
php artisan optimize

Docker users: Pull the latest image and restart your containers.