Code
Hub
Workspaces
Following
Trending
Connect
MCP
copy
Create free account
hub
/
github.com/HKUDS/OpenSpace
/ functions
Functions
2,195 in github.com/HKUDS/OpenSpace
⨍
Functions
2,195
◇
Types & classes
373
↳
Endpoints
35
↓ 2 callers
Method
wireSearch
()
showcase/my-daily-monitor/src/components/CodeStatusPanel.ts:122
↓ 2 callers
Method
wireWatchInput
(patterns: string[])
showcase/my-daily-monitor/src/components/DevOpsPanel.ts:585
↓ 1 callers
Method
__init__
Initialize ToolRanker. Args: model_name: Embedding model name. If None, will use env or config value. cache_d
openspace/grounding/core/search_tools.py:38
↓ 1 callers
Method
__init__
Initialize the wrapper. Args: wrapped_stream: The underlying stderr stream
openspace/grounding/backends/mcp/transport/task_managers/stdio.py:32
↓ 1 callers
Method
__init__
( self, *, session_id: str, config: SessionConfig, deep_research_api_k
openspace/grounding/backends/web/session.py:83
↓ 1 callers
Method
__init__
(self, record_details: bool = True)
gdpval_bench/token_tracker.py:192
↓ 1 callers
Method
_acquire_adapter_locks
(self)
openspace/communication/gateway.py:194
↓ 1 callers
Method
_add_tool_result
Add tool result to message history. IMPORTANT: Put screenshot BEFORE text for consistency with initial message.
openspace/grounding/backends/gui/anthropic_client.py:403
↓ 1 callers
Method
_after_connect
Hook called after connection is established. Override in subclasses if needed.
openspace/grounding/core/transport/connectors/base.py:79
↓ 1 callers
Function
_append_no_proxy_hosts
(*hosts: str)
openspace/communication/gateway.py:32
↓ 1 callers
Method
_append_to_file
Append action to JSONL file.
openspace/recording/action_recorder.py:157
↓ 1 callers
Method
_append_to_traj_file
Add step to traj.jsonl file
openspace/recording/recorder.py:182
↓ 1 callers
Function
_apply_env_overrides
(config: CommunicationConfig)
openspace/communication/config.py:244
↓ 1 callers
Function
_apply_replacements
Apply pre-sorted replacements in reverse order to avoid index shift.
openspace/skill_engine/patch.py:740
↓ 1 callers
Method
_arun
(self, **kwargs)
openspace/grounding/core/tool/local_tool.py:28
↓ 1 callers
Function
_assemble_essential_only
Fallback: even essential content exceeds budget. Keep: - User instruction (priority 0) — never truncated - Final iteration (priority
openspace/skill_engine/conversation_formatter.py:247
↓ 1 callers
Function
_assemble_with_budget
Assemble segments into final text respecting the character budget. Strategy: 1. Include all segments with priority ≤ 3 (CRITICAL + HIGH) in
openspace/skill_engine/conversation_formatter.py:187
↓ 1 callers
Method
_auto_record_execution
Auto-record tool execution to recording manager and quality manager.
openspace/grounding/core/tool/base.py:235
↓ 1 callers
Method
_auto_record_tool_results
Record tool execution results from LLM complete() (Path B, aligned with AnyTool).
openspace/recording/manager.py:685
↓ 1 callers
Function
_backup_skill_db
Copy the current .openspace/openspace.db to dest.
gdpval_bench/run_benchmark.py:1133
↓ 1 callers
Method
_before_connect
Hook called before establishing connection. Override in subclasses if needed.
openspace/grounding/core/transport/connectors/base.py:75
↓ 1 callers
Method
_before_connect
Negotiate transport type and set up the appropriate connection manager. Tries transports in order: 1. Streamable HTTP (new MC
openspace/grounding/backends/mcp/transport/connectors/http.py:140
↓ 1 callers
Method
_before_disconnect
Hook called before disconnection. Override in subclasses if needed.
openspace/grounding/core/transport/connectors/base.py:92
↓ 1 callers
Method
_before_disconnect
Clean up resources before disconnection.
openspace/grounding/backends/mcp/transport/connectors/http.py:357
↓ 1 callers
Method
_bm25_phase
BM25 rough-rank to keep top candidates for embedding stage.
openspace/cloud/search.py:140
↓ 1 callers
Function
_bool_arg
(name: str, default: bool)
openspace/dashboard_server.py:264
↓ 1 callers
Method
_build_adapters
(self)
openspace/communication/gateway.py:160
↓ 1 callers
Method
_build_analysis_prompt
Build the LLM prompt for execution analysis. ``context["selected_skills"]`` contains true ``skill_id`` values.
openspace/skill_engine/analyzer.py:508
↓ 1 callers
Method
_build_context_from_analysis
Build EvolutionContext from a single analysis suggestion. Loads all target skills referenced by ``suggestion.target_skill_ids``. For
openspace/skill_engine/evolver.py:1426
↓ 1 callers
Method
_build_final_result
Build final execution result. Args: instruction: Original instruction messages: Complete conversatio
openspace/agents/grounding_agent.py:793
↓ 1 callers
Function
_build_parser
()
openspace/communication/gateway.py:507
↓ 1 callers
Method
_build_planning_prompt
Build prompt for LLM planning. Args: task_description: The task to accomplish action_history: Previo
openspace/grounding/backends/gui/tool.py:397
↓ 1 callers
Function
_build_session_key_hint
(source: ChannelSource)
openspace/communication/adapters/feishu.py:826
↓ 1 callers
Function
_build_session_key_hint
(source: ChannelSource)
openspace/communication/adapters/whatsapp.py:452
↓ 1 callers
Method
_build_skill_selection_prompt
Build the prompt for LLM skill selection. Uses a plan-then-select pattern: the LLM first writes a brief execution plan, then selects
openspace/skill_engine/registry.py:666
↓ 1 callers
Function
_build_sse_candidate_urls
Try the common FastMCP `/sse` endpoint before the raw base URL.
openspace/grounding/backends/mcp/transport/connectors/http.py:33
↓ 1 callers
Function
_build_timeline
(actions: List[Dict[str, Any]], trajectory: List[Dict[str, Any]])
openspace/dashboard_server.py:552
↓ 1 callers
Method
_build_tool_status_map
Build {tool_key: has_any_success} from raw traj records. Used for deduplication: if all calls for a tool already failed (rule-based c
openspace/skill_engine/analyzer.py:250
↓ 1 callers
Function
_build_workflow_artifacts
(workflow_dir: Path, workflow_id: str, metadata: Dict[str, Any])
openspace/dashboard_server.py:582
↓ 1 callers
Function
_calc_agent_subset_token_cost
Calculate token cost for subset tasks from token_costs.jsonl per-task records.
gdpval_bench/calc_subset_performance.py:100
↓ 1 callers
Method
_capture_screenshot
Capture screenshot automatically through platforms.ScreenshotClient
openspace/recording/recorder.py:147
↓ 1 callers
Function
_check_environment
Pre-flight checks: API keys, dependencies, data availability.
gdpval_bench/run_benchmark.py:1351
↓ 1 callers
Function
_check_health
(config_path: Optional[str], host: Optional[str], port: Optional[int])
openspace/communication/gateway.py:554
↓ 1 callers
Method
_check_npm_package
Check if the npm package is globally installed. Args: package_name: npm package name Returns:
openspace/grounding/backends/mcp/installer.py:256
↓ 1 callers
Method
_check_package_installed
Check if the package is installed. Args: command: The command to check (e.g. "npx", "uvx") args: The argument
openspace/grounding/backends/mcp/installer.py:117
↓ 1 callers
Method
_check_python_package
Check if the Python package is installed as a uvx tool. uvx tools are installed in ~/.local/share/uv/tools/ directory, not in
openspace/grounding/backends/mcp/installer.py:279
↓ 1 callers
Method
_check_server_availability
Check if local server is available
openspace/recording/manager.py:539
↓ 1 callers
Method
_check_task_completion
(self, messages: List[Dict])
openspace/agents/grounding_agent.py:913
↓ 1 callers
Method
_check_task_status
(self, response: str, execution_result: str, last_error: str)
openspace/grounding/backends/shell/session.py:681
↓ 1 callers
Method
_check_uv_pip_package
Check if a Python package is installed via uv pip. Args: package_name: Python package name Returns:
openspace/grounding/backends/mcp/installer.py:326
↓ 1 callers
Method
_check_webhook_rate_limit
(self, rate_key: str)
openspace/communication/adapters/feishu.py:714
↓ 1 callers
Method
_check_workspace_artifacts
Check workspace directory for existing artifacts that might be relevant to the task. Enhanced to detect if task might already be comp
openspace/agents/grounding_agent.py:651
↓ 1 callers
Method
_cleanup_on_connect_failure
Hook called when connection fails. Override in subclasses if needed.
openspace/grounding/core/transport/connectors/base.py:83
↓ 1 callers
Method
_cleanup_on_connect_failure
Override to add MCP-specific cleanup on connection failure.
openspace/grounding/backends/mcp/transport/connectors/base.py:113
↓ 1 callers
Method
_cleanup_screenshot_client
Cleanup screenshot client resources
openspace/recording/recorder.py:255
↓ 1 callers
Method
_cleanup_wal_on_startup
Remove stale WAL/SHM left by unclean shutdown. If the main DB file is empty (0 bytes) but WAL/SHM companions exist, the database is u
openspace/skill_engine/store.py:240
↓ 1 callers
Method
_clear_webhook_anomaly
(self, remote_ip: str)
openspace/communication/adapters/feishu.py:746
↓ 1 callers
Function
_client_ip_from_request
(request: web.Request)
openspace/communication/adapters/feishu.py:766
↓ 1 callers
Method
_close_connection
Close the connection. This method should be implemented by subclasses to close the specific type of connection.
openspace/grounding/core/transport/task_managers/base.py:49
↓ 1 callers
Function
_cloud_search_and_import
Search cloud for skills relevant to *task* and auto-import top hits. This is **stage 1** of a two-stage pipeline: Stage 1 (here): server-si
openspace/mcp_server.py:373
↓ 1 callers
Function
_collect_iteration_segments
Extract segments from a ``type: "iteration"`` conversation entry. Key design decisions: - Tool calls and tool errors share the SAME high pr
openspace/skill_engine/conversation_formatter.py:87
↓ 1 callers
Function
_collect_setup_segments
Extract segments from a ``type: "setup"`` conversation entry. Only the user instruction is extracted. System prompts (including skill inject
openspace/skill_engine/conversation_formatter.py:62
↓ 1 callers
Method
_collect_text_files
Collect text files as ``{relative_path: content}``.
openspace/cloud/client.py:442
↓ 1 callers
Method
_compute_content_diff
Compute content_diff for the upload. - public + single parent → diff vs ancestor - public + no parent → add-all diff - else →
openspace/cloud/client.py:503
↓ 1 callers
Function
_compute_files_diff
Compute combined unified diff from two snapshot dicts.
openspace/skill_engine/patch.py:944
↓ 1 callers
Function
_compute_replacements
Compute (start, old_count, new_lines) replacement tuples.
openspace/skill_engine/patch.py:682
↓ 1 callers
Function
_configure_ollama_process_env
(model: str)
openspace/communication/gateway.py:45
↓ 1 callers
Method
_connection_task
Run the connection task. This task establishes and maintains the connection until cancelled.
openspace/grounding/core/transport/task_managers/base.py:162
↓ 1 callers
Function
_cosine_similarity
Compute cosine similarity between two vectors.
openspace/skill_engine/skill_ranker.py:501
↓ 1 callers
Function
_create_argument_parser
Create command-line argument parser
openspace/__main__.py:142
↓ 1 callers
Method
_dedup_and_limit
Deduplicate by name and apply limit.
openspace/cloud/search.py:220
↓ 1 callers
Function
_deep_merge_dict
Deep merge two dictionaries, update's values will override base's values
openspace/config/loader.py:26
↓ 1 callers
Function
_deep_sanitize
Recursively sanitize a JSON schema to conform to JSON Schema draft 2020-12. Removes non-standard fields and ensures valid structure.
openspace/grounding/backends/mcp/tool_converter.py:54
↓ 1 callers
Method
_default_system_prompt
Default system prompt tailored to the agent's actual backend scope.
openspace/agents/grounding_agent.py:424
↓ 1 callers
Method
_detect_screen_device
Return the screen device number of avfoundation, like '1:none' On macOS, ffmpeg -f avfoundation -list_devices true -i "" wil
openspace/local_server/platform_adapters/macos_adapter.py:401
↓ 1 callers
Method
_diagnose_skill_health
Diagnose what type of evolution a skill needs based on metrics. Returns (None, "") if the skill appears healthy. Thresholds are inten
openspace/skill_engine/evolver.py:1564
↓ 1 callers
Function
_dict_to_tool_result
Convert ClawWork tool dict to OpenSpace ToolResult.
openspace/grounding/backends/shell/productivity_tools.py:60
↓ 1 callers
Function
_discover_artifacts
Discover agent-created artifacts in the workspace. Scans ``workspace_dir`` for files matching ``_ARTIFACT_EXTENSIONS`` that are NOT the downl
gdpval_bench/run_benchmark.py:192
↓ 1 callers
Method
_dispatch_run
(self, **kwargs)
openspace/grounding/core/tool/local_tool.py:14
↓ 1 callers
Function
_download_via_curl
Download using curl subprocess — bypasses Python SSL entirely.
gdpval_bench/task_loader.py:661
↓ 1 callers
Function
_download_via_requests
Download using requests library with retry adapter.
gdpval_bench/task_loader.py:713
↓ 1 callers
Function
_download_via_urllib
Download using urllib with relaxed SSL context (last resort).
gdpval_bench/task_loader.py:754
↓ 1 callers
Function
_download_via_wget
Download using wget subprocess — bypasses Python SSL entirely.
gdpval_bench/task_loader.py:688
↓ 1 callers
Function
_edit_distance
Levenshtein edit distance (compact DP, O(min(m,n)) space).
openspace/skill_engine/analyzer.py:109
↓ 1 callers
Method
_enhance_result
Enhance tool result with LLM-based visual analysis.
openspace/agents/visual_analyzer.py:88
↓ 1 callers
Function
_enrich_with_pricing
Merge pricing data from task_values.jsonl into tasks that lack it.
gdpval_bench/task_loader.py:331
↓ 1 callers
Method
_ensure_dependencies_impl
Internal implementation of ensure_dependencies (called within lock).
openspace/grounding/backends/mcp/installer.py:564
↓ 1 callers
Function
_ensure_local_no_proxy
()
openspace/host_detection/resolver.py:78
↓ 1 callers
Method
_ensure_model
Ensure embedding model is ready (local or remote).
openspace/grounding/core/search_tools.py:256
↓ 1 callers
Method
_establish_connection
Establish the connection. This method should be implemented by subclasses to establish the specific type of connection needed.
openspace/grounding/core/transport/task_managers/base.py:34
↓ 1 callers
Function
_evaluate_task
Evaluate agent artifacts using ClawWork's LLMEvaluator. Fully aligned with ClawWork's evaluation pipeline: 1. Discover artifacts in workspa
gdpval_bench/run_benchmark.py:271
↓ 1 callers
Method
_evict_idle_loop
(self)
openspace/communication/runtime_manager.py:138
↓ 1 callers
Method
_evolve_captured
Capture a novel pattern as a brand-new skill. Uses agent loop for information gathering + apply-retry cycle.
openspace/skill_engine/evolver.py:928
↓ 1 callers
Method
_evolve_derived
Create enhanced version in a new directory. Supports single-parent (enhance) and multi-parent (merge/fuse). Uses agent loop for infor
openspace/skill_engine/evolver.py:796
↓ 1 callers
Method
_evolve_fix
In-place fix: same name, same directory, new version record. Uses agent loop for information gathering + apply-retry cycle.
openspace/skill_engine/evolver.py:702
↓ 1 callers
Method
_execute_code_from_response
execute the code and track the code block Returns: Tuple[Optional[Dict], str]: (code_info, execution_result)
openspace/grounding/backends/shell/session.py:581
↓ 1 callers
Method
_execute_planned_action
Execute a planned action through the connector. Args: action: Action dict with action_type and parameters
openspace/grounding/backends/gui/tool.py:477
↓ 1 callers
Method
_execute_task_with_planning
Execute task with LLM-based planning loop. Planning loop: 1. Observe: Get screenshot 2. Plan: LLM decides ne
openspace/grounding/backends/gui/tool.py:123
↓ 1 callers
Function
_execute_tool_call
Execute LLMClient returned tool_call Args: tool: BaseTool instance (must be obtained from GroundingClient and bound to runtime_info)
openspace/llm/client.py:344
↓ 1 callers
Function
_extract_embedded_summary
Extract self-generated summary from tool result content. Shell agent results often contain an ``Execution Summary (N steps):`` block that pro
openspace/skill_engine/conversation_formatter.py:313
← previous
next →
601–700 of 2,195, ranked by callers