Draft
Conversation
- FleetTaskEnv wraps FleetEnvClient with task-oriented interface - Accepts task configs from export_training_tasks.py - Creates versioned environments on reset - Injects task prompt into observations - Executes verifier for reward computation on episode completion - Supports both sync and async step methods - Factory functions: make_fleet_task_env, from_json_file - Tests: 20 unit tests for init, specs, verifiers, factories 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
The MCP images don't exist for all environment versions, causing FleetVersionNotFoundError when trying to create environments. Changing the default to None allows the Fleet SDK to use standard images which are available for all versions. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
FleetEnvClient.from_fleet() was not accepting data_key/data_version parameters, causing them to be passed through **kwargs to HTTPEnvClient which doesn't accept them. - Add data_key and data_version as explicit parameters - Pass them to fleet.make() - Update task_env.py to pass them separately 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Fleet SDK expects data_key in "key:version" format, not as separate parameters. Updated from_fleet() to combine them before calling fleet.make(). 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
HTTPEnvClient.reset() doesn't support seed parameter yet. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Increases default timeout from 15s to 60s for Fleet API calls. This prevents timeouts during environment initialization. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Previously reset() did partial work and reset_async() added tool fetching. Now reset_async() does all the work (including fetching tools) and reset() is just a sync wrapper that calls it via run_until_complete(). This ensures both methods return identical results including tools. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
MCP's call_tool() returns a CallToolResult Pydantic object, not plain text. This was causing ugly repr strings to be passed to agents like: "meta=None content=[TextContent(type='text', text='...')] ..." Now properly extracts: - Text content from result.content[].text - Tries JSON parsing for structured results - Falls back to structuredContent if available - Handles isError cases 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Tests for: - FleetMCPClient._extract_tool_result(): - Single text content extraction - JSON parsing from text - Multiple text contents - Error result handling - Structured content fallback - Empty result handling - FleetTaskEnv reset: - reset_async() returns tools - reset() calls reset_async() (sync wrapper) 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Move fleet.make() and list_tools() into FleetTaskEnv.__init__() - Tools are now fetched at env creation, not during reset - reset_async() calls _orch.reset() with error handling, returns cached tools - Use asyncio.run() for Python 3.13 compatibility - Update tests for new initialization pattern 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Log task_key and verifier code preview when verifier fails - Catch syntax errors separately with clear message - Show which functions were found if 'verify' is missing Helps debug issues like "Verifier code must define a 'verify' function" 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Replace custom _execute_verifier_local() with Fleet SDK's Task.verify_detailed() which properly sets up the verifier namespace with: - Environment type annotation - Helper functions (normalized_contains, etc.) - Proper function discovery (not just "verify" function) This fixes "name 'Environment' is not defined" errors during verifier execution. Changes: - _compute_reward: Create Fleet SDK Task and call verify_detailed() - Support both 'verifier_code' and 'verifier_func' field names - Add comprehensive logging for debugging - Remove broken _execute_verifier_local method Tests: - Update all verifier tests to mock Fleet SDK Task.verify_detailed() - Add tests for various edge cases (no verifier, no orch, exceptions) - Fix fixture to avoid asyncio.run() conflicts with pytest-asyncio 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
8 tasks
- Add retry with exponential backoff (3 attempts, 1s/2s/4s delays) - Log errors instead of silently swallowing exceptions - Log warning when some clients fail but others succeed - Log error after all retries exhausted This fixes silent failures when MCP connections are flaky, which caused 'no tools found' errors in SkyRL training.
call_tool now retries with exponential backoff (3 attempts, 1s/2s/4s) on connection errors, similar to list_tools. ValueError (tool not found) is not retried. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Adds exponential backoff retry (3 attempts, 2s base delay) around fleet.make() to handle transient Fleet API errors like health check failures that can occur during instance provisioning. Only retries on transient errors (health check, timeout, connection). Permanent errors are raised immediately. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Add Toolathlon-style context management tools for long trajectories: - check_context: Check visible/total turn counts - manage_context: Drop old turns to free up context space - search_history: Search all history (including dropped) - search_tool_output: Search truncated tool output - view_tool_output: Paginate through truncated output The ContextManager class can be used by any training framework that maintains chat_history. It tracks full history and handles truncated tool outputs. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Computer-use tasks require MCP-enabled container images (e.g., famazon:mcp0.0.7) which have scrot installed for screenshots and the MCP server with 'computer' tool for mouse/keyboard control.
Previously, tools were only fetched for tool_use modality due to a restrictive condition. This caused computer_use tasks to fail with "no tools found in observation" because the computer tool (mouse, keyboard, screenshot) was never fetched. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
When task_modality is computer_use, filter tools to only include the 'computer' tool. This prevents the model from using API tools when it should be using mouse/keyboard control. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Two critical fixes for VL (vision-language) model training: 1. ImageContent extraction: _extract_tool_result() now handles MCP ImageContent (base64 images with mimeType) and converts them to OpenAI-compatible format for VL models. 2. Tool filtering: computer_use modality now always filters to only the 'computer' tool. If no computer tool found, clears all tools and logs warning (prevents model from using API tools). Tests added: - test_extract_image_content - test_extract_mixed_text_and_image_content - test_extract_image_default_mimetype - test_computer_use_filters_to_computer_tool - test_computer_use_clears_tools_when_no_computer_tool - test_tool_use_does_not_filter - test_computer_use_filters_function_format 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
For VL (vision-language) models doing computer_use tasks, the model needs visual input to know where to click. Previously, reset() only returned metadata without a screenshot, leaving VL models blind. Now for computer_use modality, reset_async() automatically takes a screenshot after reset and includes it in the observation as `initial_screenshot`. This is in OpenAI-compatible format for VL models. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
The manager API (POST /reset) hangs indefinitely on some env images (e.g. google-maps v0.0.53). Since reset failure is already handled gracefully (warning + continue), this adds a short dedicated timeout (default 10s) so the reset fails fast instead of blocking for the full request_timeout_s (60-120s). This saves 50-110s per episode during training when the manager API is unresponsive, while still allowing reset to succeed on healthy envs. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Add env_key property to FleetTaskEnv - Prefix all error/warning logs with [env=X] for easy filtering - Helps identify which environments have infrastructure issues (502s, health checks) 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Defense-in-depth against zombie threads: if asyncio cancellation somehow fails to propagate, HTTP-level timeouts ensure MCP calls fail within 2 minutes instead of hanging forever. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- mcp_tools.py: raise RuntimeError after 3 failed list_tools attempts instead of silently returning empty ListToolsAction - mcp_tools.py: increase retry_base_delay from 1s to 2s - task_env.py: don't set _tools_fetched=True on failure so next reset_async() can retry tool discovery Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
dzorlu
force-pushed
the
deniz/fleet_client
branch
2 times, most recently
from
272b0b3 to
8370cd1
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
1 participant
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.
PR: Fleet environments (OpenEnv)
This PR documents and refines the Fleet runtime integration for OpenEnv.
What this enables
reset / step / statetools/list + tools/callWhat this is not
Main abstractions
FleetEnvClient(HTTP): orchestrator handle for reset/step/state.FleetMCPTools(MCP): agent handle for listing/calling tools.api/v1/mcpandmcp)convert_tool_format)Quickstart
pip install "openenv-core[fleet]"export FLEET_API_KEY="..."python examples/fleet_env_example.py <env_key>References
rfcs/001-abstractions.mdrfcs/003-mcp-support.mdTODOs / known sharp edges
api/v1/mcpvsmcp)