Jules
Google's async AI coding agent with proactive automation: Suggested Tasks scans repos for improvements, Scheduled Tasks automates recurring work, CI Fixer auto-resolves failing builds, and Render integration closes the deployment failure loop.
Dimension breakdown
Score · confidenceGoogle's async AI coding agent with proactive automation: Suggested Tasks scans repos for improvements, Scheduled Tasks automates recurring work, CI Fixer auto-resolves failing builds, and Render integration closes the deployment failure loop. Gemini 3.1 Pro powers the agent (default for paid tiers since March 2026); Jules-the-product scores ~51.8% on SWE-bench Verified — materially below the 80.6% raw-model figure and below peer agents like Claude Code (~80.8%).
MCP support (Feb 2026) adds Linear/Neon/Supabase/Context7 connectors. Jitro / Jules V2 (goal-oriented agentic workspace) did NOT ship at Google I/O 2026 (May 19) — the developer keynote led with Gemini 3.5 and Antigravity 2.0, and Jitro remains pre-launch/waitlist.
Enterprise auth still absent — paid plans are individual Google Accounts only, no Workspace/SSO/audit logs in official docs; invisible-Unicode prompt injection (May 2025 disclosure) still without confirmed public remediation 12+ months on.
GitHub-only SCM — no GitLab/Bitbucket.
Use cases
Not yet assessed — this section fills in as ACES research covers the tool.
Risk flags
No active caps — no risk flags apply to this tool right now.
Status rationale
Tracked maintained. Strong proactive agent capabilities (Suggested Tasks, Scheduled Tasks, Render integration, CI Fixer, Planning Critic) on Gemini 3.1 Pro with genuine async workflow value, plus a Feb 2026 MCP connector ecosystem (Linear/Neon/Supabase/Context7).
Rating 72. Jitro/V2 (goal-oriented workspace) did NOT ship at Google I/O 2026 (May 19) — the developer keynote centered on Gemini 3.5 and Antigravity 2.0; Jitro remains pre-launch/waitlist.
Blocked from Assessed by: no confirmed enterprise auth in official docs (SSO/SAML/audit logs), no Jules-specific compliance certs, no Workspace account support confirmed (paid plans are individual Google Accounts only), unresolved May 2025 invisible-Unicode prompt-injection research disclosure, GitHub-only SCM, and a vendor-reported product SWE-bench Verified score (~51.8%) that materially trails peer agents.
Movement triggers
Upgrade to Assessed if: (1) Jitro/Jules V2 actually ships from its post-I/O waitlist with confirmed enterprise auth (Workspace SSO/SAML, audit logs) and a Jules-specific compliance certification (I/O 2026 came and went on May 19 without a Jitro GA reveal — it is now a post-keynote waitlist watch item, not a keynote bet; [2026-06-28: verify Jitro status — may be past waitlist into open testing per June 2026 sources]); (2) named enterprise customer deployment confirmed in official Google sources; (3) prompt injection research disclosure receives official remediation announcement; (4) GitLab or Bitbucket SCM support ships; (5) independent (non-vendor) SWE-bench / agent benchmark confirms materially higher product reliability. Downgrade to Detected if: prompt injection exploited in a documented production incident with customer impact; Jules v1 deprecation announced without enterprise-ready Jitro replacement; community reports of major reliability regression after Jitro transition.
Risks & limitations
Not yet assessed — this section fills in as ACES research covers the tool.
Integration surface
Not yet assessed — this section fills in as ACES research covers the tool.
Adoption & benchmarks
Not yet assessed — this section fills in as ACES research covers the tool.
Spotted something wrong or missing here? Suggest a change →
Per-source contributions
Click any dimension to see the underlying sources and citations.
More in this category