Add OpenHands Cloud API automation scripts and workflow#103
Closed
Add OpenHands Cloud API automation scripts and workflow#103
Conversation
Co-authored-by: openhands <openhands@all-hands.dev> Co-authored-by: Engel Nyst <enyst@users.noreply.github.com>
…nds#10493) Co-authored-by: openhands <openhands@all-hands.dev>
Co-authored-by: Engel Nyst <enyst@users.noreply.github.com>
…penHands#10412) Co-authored-by: openhands <openhands@all-hands.dev>
Co-authored-by: openhands <openhands@all-hands.dev>
…penHands#10537) Co-authored-by: openhands <openhands@all-hands.dev>
…warning instead of error (OpenHands#10502)
…8.18 (OpenHands#10561) Co-authored-by: openhands <openhands@all-hands.dev>
…penHands#9707) Co-authored-by: Engel Nyst <enyst@users.noreply.github.com>
…; apply across entrypoints; add tests (OpenHands#10540) Co-authored-by: openhands <openhands@all-hands.dev>
Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: openhands <openhands@all-hands.dev> Co-authored-by: Graham Neubig <neubig@gmail.com> Co-authored-by: llamantino <213239228+llamantino@users.noreply.github.com> Co-authored-by: mamoodi <mamoodiha@gmail.com> Co-authored-by: Tim O'Farrell <tofarr@gmail.com> Co-authored-by: Hiep Le <69354317+hieptl@users.noreply.github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: Ryan H. Tran <descience.thh10@gmail.com> Co-authored-by: Neeraj Panwar <49247372+npneeraj@users.noreply.github.com> Co-authored-by: sp.wack <83104063+amanape@users.noreply.github.com> Co-authored-by: Insop <1240382+insop@users.noreply.github.com> Co-authored-by: test <test@test.com> Co-authored-by: Engel Nyst <enyst@users.noreply.github.com> Co-authored-by: Zhonghao Jiang <zhonghao.J@outlook.com> Co-authored-by: Ray Myers <ray.myers@gmail.com>
…gainst data exfiltration (OpenHands#10477) Co-authored-by: Engel Nyst <enyst@users.noreply.github.com> Co-authored-by: openhands <openhands@all-hands.dev>
Co-authored-by: Xingyao Wang <xingyao@all-hands.dev>
Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: openhands <openhands@all-hands.dev> Co-authored-by: Xingyao Wang <xingyao@all-hands.dev> Co-authored-by: Graham Neubig <neubig@gmail.com> Co-authored-by: llamantino <213239228+llamantino@users.noreply.github.com> Co-authored-by: mamoodi <mamoodiha@gmail.com> Co-authored-by: Tim O'Farrell <tofarr@gmail.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: Ryan H. Tran <descience.thh10@gmail.com> Co-authored-by: Neeraj Panwar <49247372+npneeraj@users.noreply.github.com> Co-authored-by: sp.wack <83104063+amanape@users.noreply.github.com> Co-authored-by: Insop <1240382+insop@users.noreply.github.com> Co-authored-by: test <test@test.com> Co-authored-by: Engel Nyst <enyst@users.noreply.github.com> Co-authored-by: Zhonghao Jiang <zhonghao.J@outlook.com> Co-authored-by: Ray Myers <ray.myers@gmail.com>
…tests (OpenHands#10538) Co-authored-by: openhands <openhands@all-hands.dev>
…nt (OpenHands#10579) Co-authored-by: openhands <openhands@all-hands.dev>
…handler (OpenHands#10403) Co-authored-by: openhands <openhands@all-hands.dev>
…0 in /frontend in the eslint group (OpenHands#10511) Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
…management, memory UI) (OpenHands#10519) Co-authored-by: amanape <83104063+amanape@users.noreply.github.com> Co-authored-by: openhands <openhands@all-hands.dev>
…ands#10595) Co-authored-by: amanape <83104063+amanape@users.noreply.github.com> Co-authored-by: openhands <openhands@all-hands.dev>
…agent management) (OpenHands#10596) Co-authored-by: amanape <83104063+amanape@users.noreply.github.com> Co-authored-by: openhands <openhands@all-hands.dev>
Signed-off-by: hereisok <hereisok@angai.wk@gmail.com> Co-authored-by: mamoodi <mamoodiha@gmail.com>
Co-authored-by: openhands <openhands@all-hands.dev>
Co-authored-by: OpenHands <openhands@all-hands.dev>
Features added: - Dropdown in GitHub Actions to select from available prompt files - Text field for custom prompts that override prompt files - New example prompt files: debug_issue.j2, code_review.j2, feature_implementation.j2 - Enhanced CLI with --prompt-file and --custom-prompt options - Custom prompts take precedence over prompt files - Improved user experience with clear feedback on prompt selection The workflow now supports: 1. Selecting from predefined prompt templates via dropdown 2. Entering custom prompts directly in the workflow interface 3. Fallback to default new_conversation.j2 if no prompt specified Co-authored-by: openhands <openhands@all-hands.dev>
Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com>
Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com>
Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com>
Major fixes: - Remove hard-coded GitHub repo/issue targets (security risk) - Add --comment-repo and --comment-issue CLI parameters for safe commenting - Update function signature to accept comment parameters - Remove GITHUB_TOKEN from workflow (no longer needed without hard-coded targets) - Remove issues: write permission (no longer needed) - Fix hard-coded conversation URL to use client.base_url - Replace eval with safe array-based command building - Add working-directory to workflow to avoid cd All commenting is now opt-in and parameterized, eliminating security risks. Tests still pass and functionality is preserved.
- Add back GITHUB_TOKEN and issues: write permission - Automatically detect and comment on: - Current PR when triggered from pull_request event - Current issue when triggered from issues event - Issue 95 when triggered manually via workflow_dispatch - Uses github.repository and github.event context for dynamic targeting - Maintains security by using GitHub's built-in context instead of hard-coding This restores the desired automatic posting behavior while being smarter about which PR/issue to comment on based on the trigger context.
Issue 95 is the dedicated conversation log issue, so always post success/error comments there regardless of trigger context. This keeps all conversation results in one place for easy tracking.
…mon_tail.j2 Co-authored-by: openhands <openhands@all-hands.dev>
Co-authored-by: openhands <openhands@all-hands.dev>
- create_conversation.py: Enhanced conversation creation with multi-stage polling and LLM model detection - agent_conversations.md: Tracking file for 17 active conversations with status and LLM models - scripts/cloud_api.py: Enhanced API client with get_events method and condensation event examples - scripts/prompts/common_tail.j2: Added shell safety warnings for agent runtime sessions - todo_later.md: Future enhancement ideas including Stop prompt functionality This framework enables automated management of multiple OpenHands Cloud conversations with comprehensive tracking, LLM model detection, and event querying capabilities.
## Model Switching Discovery in Issue OpenHands#10730 Conversation **UNPRECEDENTED FINDING**: OpenHands Cloud automatically switched LLM models mid-conversation during Issue OpenHands#10730 analysis. ### Key Discovery: - **Conversation ID**: 5e1c2e15da9842d196ae29e44fc45f36 - **Total Events**: 297 events with 148 tool calls - **Model Switch**: GPT-5 → Claude Sonnet 4 at event 252→262 ### Model Usage Breakdown: - **GPT-5 Phase**: 134 tool calls (events 9-252) - Primary work phase - analysis and implementation - Model: litellm_proxy/gpt-5-2025-08-07 - **Claude Sonnet 4 Phase**: 14 tool calls (events 262-296) - Final phase - completion and refinement - Model: litellm_proxy/claude-sonnet-4-20250514 ### Significance: - **First documented case** of automatic model switching in OpenHands Cloud - **Clean handoff** with 9-event gap, no overlap - **Task-appropriate routing**: GPT-5 for heavy analysis, Claude for completion - **Platform intelligence**: Demonstrates sophisticated AI orchestration ### Additional Research Findings: - **API vs UI Routing**: Different model selection algorithms - **API Deterministic**: 100% consistent content-based routing - **UI Probabilistic**: Variable routing with user behavior influence - **GPT-5 Breakthrough**: First success via UI after 12 attempts ### Files Added: - agent_conversations.md: Complete conversation fleet tracking (37 conversations) - issue_10730_complete_trajectory_297_events.json: Full 297-event trajectory (1.8MB) This discovery reveals OpenHands Cloud as a sophisticated AI orchestration platform with intelligent model routing capabilities. Co-authored-by: openhands <openhands@all-hands.dev>
…-2025-08-07)\n\nCo-authored-by: openhands <openhands@all-hands.dev>
…b6a09ada34259 (prefix discrepancy)\n\nCo-authored-by: openhands <openhands@all-hands.dev>
…oints - Add list_conversations and helpers in scripts/cloud_api.py - Append missing conversations with minimal info only (no trajectories) Co-authored-by: openhands <openhands@all-hands.dev>
…om list endpoint\n\n- get_early_model() using first 20 events only for unknowns\n- Add notes on events count and model probe strategy\n- Single-pass enrich MD with trigger and PR references\n\nCo-authored-by: openhands <openhands@all-hands.dev>
…probe (20 events)\n\nCo-authored-by: openhands <openhands@all-hands.dev>
- Add validation to clamp limit parameter between 1 and 100 - Prevents server-side errors from out-of-range limit values - Addresses review comment on PR #94 Co-authored-by: openhands <openhands@all-hands.dev>
…ons.md\n\nWe do not commit trajectory files; remove stray references.\n\nCo-authored-by: openhands <openhands@all-hands.dev>
…ls via reverse-window probe\n\nCo-authored-by: openhands <openhands@all-hands.dev>
…cy (GPT-5 without litellm_proxy/)\n\nCo-authored-by: openhands <openhands@all-hands.dev>
…itching case confirmed\n\nCo-authored-by: openhands <openhands@all-hands.dev>
Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com>
… openhands <openhands@all-hands.dev>
…l-Hands-AI/agent-sdk\n\n- Include link and initial status\n- Will populate model when detected\n\nCo-authored-by: openhands <openhands@all-hands.dev>
…d7ca4a98445d1b9e9d0fbe3fc2898\n\nCo-authored-by: openhands <openhands@all-hands.dev>
enyst
force-pushed
the
test-cloud-api
branch
from
31e6302 to
19d3c09
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
20 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This PR adds a minimal, useful toolkit to automate OpenHands Cloud via API and a GitHub Actions workflow to trigger new conversations.
What’s included
Secrets / env
Quick usage
Notes
Co-authored-by: openhands openhands@all-hands.dev
@enyst can click here to continue refining the PR
Summary by CodeRabbit
New Features
Documentation
Tests