Skip to content

Pull requests: cuga-project/cuga-agent

New pull request New
17 Open 308 Closed
17 Open 308 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

perf(policy): skip output-formatter LLM call for clean plain-text responses complexity: low Small, focused change — easy to review readability: good Clear PR goal and description; easy to review
#783 opened Sep 15, 2026 by mbwhite Loading…
1 task done
4
fix(cuga-lite): scope the VERIFY ungrounded rule to facts, not choices complexity: medium Moderate scope — multiple files or non-trivial logic readability: good Clear PR goal and description; easy to review
#782 opened Sep 15, 2026 by sami-marreed Contributor Loading…
5
fix(registry): stop an endpoint that rejects every argument set the same way complexity: medium Moderate scope — multiple files or non-trivial logic readability: good Clear PR goal and description; easy to review
#781 opened Sep 15, 2026 by sami-marreed Contributor Loading…
1
6
fix(cuga-lite): stop advertising filesystem helpers the executor never injects complexity: medium Moderate scope — multiple files or non-trivial logic readability: good Clear PR goal and description; easy to review
#780 opened Sep 15, 2026 by sami-marreed Contributor Loading…
1
5
feat(cuga-lite): native function-calling execution mode + step discipline complexity: high Large or risky change — needs careful review readability: good Clear PR goal and description; easy to review
#777 opened Sep 15, 2026 by segevshlomovIBM Collaborator Loading…
2 tasks done
@segevshlomovIBM
29
fix(policy): append pydantic format instructions for LiteLLM models complexity: medium Moderate scope — multiple files or non-trivial logic readability: good Clear PR goal and description; easy to review
#776 opened Sep 14, 2026 by mbwhite Loading…
1 task done
10
fix(policy): per-agent policy isolation — playbooks no longer bleed across agents (#767) complexity: high Large or risky change — needs careful review readability: good Clear PR goal and description; easy to review
#773 opened Sep 10, 2026 by offerakrabi Collaborator Loading…
1
@offerakrabi
19
feat: PR Regression Testing Oriented Changes complexity: medium Moderate scope — multiple files or non-trivial logic readability: fair Partial context or goal not obvious on first read
#759 opened Sep 5, 2026 by AnkitaNaik Collaborator Loading…
35
fix(cuga-lite): preserve computed results in final answers complexity: medium Moderate scope — multiple files or non-trivial logic readability: good Clear PR goal and description; easy to review
#754 opened Sep 3, 2026 by Tethys0 Loading…
5 tasks done
1
6
fix(supervisor): isolate sub-agent checkpoints across users complexity: high Large or risky change — needs careful review readability: good Clear PR goal and description; easy to review
#751 opened Sep 2, 2026 by sami-marreed Contributor Loading…
1 task done
1
10
fix(registry): normalize AppWorld file_system paths to canonical ~/ form (#730) complexity: medium Moderate scope — multiple files or non-trivial logic readability: good Clear PR goal and description; easy to review
#740 opened Sep 2, 2026 by Sergey-Zeltyn Collaborator Loading…
1 task done
1
5
feat(memory): add experimental workspace to the supported image complexity: high Large or risky change — needs careful review readability: good Clear PR goal and description; easy to review
#737 opened Sep 1, 2026 by illeatmyhat Collaborator • 3/3 Loading…
57
feat(memory): add manual retention controls complexity: high Large or risky change — needs careful review readability: good Clear PR goal and description; easy to review
#735 opened Sep 1, 2026 by illeatmyhat Collaborator • 2/3 Loading…
27
feat(memory): add experimental Evolve management API complexity: high Large or risky change — needs careful review readability: good Clear PR goal and description; easy to review
#736 opened Sep 1, 2026 by illeatmyhat Collaborator • 1/3 Loading…
12
fix(cuga-lite): auto-continue autonomous-mode deferrals instead of finalizing (#445) complexity: medium Moderate scope — multiple files or non-trivial logic readability: good Clear PR goal and description; easy to review
#732 opened Sep 1, 2026 by haroldship Collaborator • Draft
4 of 5 tasks
1
15
fix(code-extraction): keep only the first fenced block from reasoning blocked: evaluation Do not merge until evaluation / benchmark validation is done complexity: low Small, focused change — easy to review readability: good Clear PR goal and description; easy to review
#645 opened Aug 12, 2026 by sami-marreed Contributor Loading…
1 task done
1
3
feat(events & channels): WhatsApp as a direct channel (Meta Cloud API)
#605 opened Aug 8, 2026 by anupamamurthi Collaborator • 3/3 Loading…
2
ProTip! no:milestone will show everything without a milestone.