github desplega-ai/agent-swarm v2.14.2

40 minutes ago

Install & Run

npm:

npx @desplega.ai/agent-swarm@2.14.2
# or install globally
npm i -g @desplega.ai/agent-swarm@2.14.2

bun:

bunx @desplega.ai/agent-swarm@2.14.2
# or install globally
bun i -g @desplega.ai/agent-swarm@2.14.2

Docker:

# Server
docker pull ghcr.io/desplega-ai/agent-swarm:2.14.2
# Worker
docker pull ghcr.io/desplega-ai/agent-swarm-worker:2.14.2
# Dashboard
docker pull ghcr.io/desplega-ai/agent-swarm-ui:2.14.2

E2B templates:

# API
agent-swarm-api-2-14-2
agent-swarm-api-latest

# Worker / lead runtime
agent-swarm-worker-2-14-2
agent-swarm-worker-latest

agent-fs

Ships agent-fs 0.18.37 (bumped from 0.18.35).

agent-swarm pins agent-fs via the chart image tag and npm packages @desplega.ai/agent-fs and @desplega.ai/agent-fs-just-bash; each agent-fs release needs a chore(deps): bump agent-fs pins PR here.


What's Changed

  • fix(ui-e2e): fill the tasks search box once by @desplega-bot in #3487
  • docs(skills): systemDefault skill seed patches (2026-11-20) by @desplega-bot in #3488
  • docs: add access-permissions check to training evidence review checklist by @desplega-bot in #3489
  • docs: frame differing win-loss reasons as items for human review by @desplega-bot in #3490
  • docs: add campaign launch review guide by @desplega-bot in #3491
  • docs(guides): add near-miss investigation evidence packet guide by @desplega-bot in #3492
  • chore(docker): bump claude-code, codex, amp, gemini, anthropic sdk (2026-11-20) by @desplega-bot in #3493
  • docs(specs): sync TLA+ specs with main (heartbeat, workflows) by @desplega-bot in #3495
  • fix(tests): seed the db-query payload-cap test in one statement by @desplega-bot in #3496
  • fix(api): call this API over loopback from API-side self-calls by @desplega-bot in #3497
  • fix(workflows): refuse extension-bridged event names on the signal routes by @desplega-bot in #3486
  • fix(workflows): run edited script nodes as the creator only after the creator or a lead saves by @desplega-bot in #3494
  • chore(deps): ignore Dependabot majors and group each ecosystem into one weekly PR by @desplega-bot in #3483
  • fix: allow sheet reload rendering to finish in UI E2E by @desplega-bot in #3499
  • fix(scripts): report every line a quickjs compat finding recurs on by @desplega-bot in #3498
  • fix(workflows): route a resumed approval step on its stored request by @desplega-bot in #3485
  • fix(ui-e2e): give the object-embed sheet page loads a 15 s budget by @desplega-bot in #3500
  • chore(deps): bump agent-fs pins to 0.18.36 by @desplega-bot in #3503
  • fix(compose): bind dev MinIO ports to localhost by @desplega-bot in #3505
  • docs: add meeting decision follow-up guide by @desplega-bot in #3507
  • docs: add first-day onboarding readiness guide by @desplega-bot in #3508
  • perf(apps): scrub sync pull records with the sweep scrubber by @desplega-bot in #3506
  • chore(compose): read the MinIO login from .env and bind its ports to loopback by @desplega-bot in #3504
  • fix(rbac): declare memory.reembed on POST /api/memory/re-embed by @Seyamalam in #3502
  • docs: move Omnia use-case guides under Playbooks > Use cases by @desplega-bot in #3512
  • fix(rbac): declare task.progress.write on POST /api/tasks/{id}/progress by @Seyamalam in #3514
  • chore(docker): bump pi, amp, goose (2026-11-20) by @desplega-bot in #3516
  • fix(x402): enforce spending limits on x402 v1 maxAmountRequired by @desplega-bot in #3521
  • fix(ui): give fenced code the message column width by @tarasyarema in #3522
  • test(e2e): cover GitHub and Linear provider flows with emulate by @desplega-bot in #3519
  • fix(prompt-templates): require lead, operator or user for shared template writes by @desplega-bot in #3518
  • feat(examples): video review page, swarm bundle and docs guide by @tarasyarema in #3517
  • fix(x402): select the signed amount field by protocol version by @desplega-bot in #3523
  • test(e2e): trial @emulators/ai as the LLM backend for one scenario by @desplega-bot in #3513
  • feat(pr-body): template-declared evidence paths and Slack mock frames by @tarasyarema in #3465
  • chore(deps): bump the minor-and-patch group across 1 directory with 15 updates by @dependabot[bot] in #3501
  • docs(agents-md): slim CLAUDE.md under the Codex size cap and check the size by @tarasyarema in #3466
  • fix(docker): render emoji in Slack mock frames inside the worker image by @desplega-bot in #3524
  • fix(rbac): declare posture on every backlogged route and delete the route backlog by @desplega-bot in #3515
  • feat(pr-body): readable template with before/after evidence on every PR by @tarasyarema in #3468
  • feat(examples): share card, logo and visit counts for the video review site by @tarasyarema in #3526
  • fix(ui): send the connector code only to the exact connect page by @desplega-bot in #3525
  • feat(examples): light theme for the video review page, and a fix for its first deploy by @tarasyarema in #3527
  • fix(ui): ask before a top-level link switches the dashboard connection by @desplega-bot in #3529
  • fix(slack): list the reviewed worker's deliverables on Lead review cards by @desplega-bot in #3511
  • docs(skills): share a public page's api_url with readers outside the dashboard by @desplega-bot in #3531
  • fix(evals): refuse other-origin and non-JSON writes when the API runs without a key by @desplega-bot in #3532
  • release: v2.14.2 + agent-fs 0.18.37 + docs daily update 2026-11-21 by @desplega-bot in #3533

Full Changelog: v2.14.1...v2.14.2

Don't miss a new agent-swarm release

NewReleases is sending notifications on new releases.