MCPcopy Create free account

hub / github.com/ScaleML/AgentSPEX / functions

Functions1,694 in github.com/ScaleML/AgentSPEX

↓ 3 callersFunction_safe_print
(*args, **kwargs)
src/benchmarks/elaipbench/instance.py:17
↓ 3 callersFunction_safe_print
(*args, **kwargs)
src/benchmarks/chembench/run.py:39
↓ 3 callersFunction_save_config
(cfg: dict)
src/harness/ephemeral/repl.py:97
↓ 3 callersMethod_validate_task
Validate task data format
src/harness/parsing/yaml_parser.py:76
↓ 3 callersMethodadd_assistant_message
Add assistant message, optionally with tool calls.
src/harness/agentic_loop/session.py:44
↓ 3 callersFunctionapply_line_with_indent
Apply proper indentation to a line based on its relative indent in patch.
src/sandbox_vm/tools/code/patch.py:442
↓ 3 callersFunctionassignDefined
Add fields from `src` to `dst`, skipping undefined values.
yaml-flow-editor/src/converters/yaml-to-graph.ts:39
↓ 3 callersMethodbest_effort_host_write_text
Best-effort mirror write on host filesystem.
src/harness/checkpoints/checkpoint_manager.py:339
↓ 3 callersFunctionbroadcastStatus
(run: RunInfo)
yaml-flow-editor/vite.config.ts:274
↓ 3 callersFunctioncall_tool
(tool_name: str, **kwargs)
tests/tool_tests/test_terminal.py:15
↓ 3 callersFunctioncleanup
(signal: NodeJS.Signals | 'exit')
yaml-flow-editor/vite.config.ts:359
↓ 3 callersFunctioncreateNodeFilter
* Filter out elements that should not be exported (MiniMap, Controls, etc.)
yaml-flow-editor/src/utils/image.ts:60
↓ 3 callersFunctiondashboardProxyPath
(run: RunInfo)
yaml-flow-editor/vite.config.ts:70
↓ 3 callersMethodemit_full_state
(self, sid: str)
scripts/dashboard.py:472
↓ 3 callersMethodemit_update
(self)
scripts/dashboard.py:475
↓ 3 callersMethodexecute_call_step
Execute a sub-module workflow. Calls another YAML workflow file as a reusable sub-module with parameter passing. The sub-module runs
src/harness/execution/step_handlers/calls.py:43
↓ 3 callersMethodexecute_conditional_step
Execute if/then/else conditional branching. Evaluates a condition and executes either the 'then' or 'else' branch. Args:
src/harness/execution/step_handlers/control.py:219
↓ 3 callersMethodexecute_for_each_step
Execute a for_each loop over items. Iterates over a list, executing nested steps for each item. The current item is made available in
src/harness/execution/step_handlers/control.py:127
↓ 3 callersMethodexecute_increment_step
Increment an integer variable in context. Args: step_data: Contains 'increment' key with the variable name. context:
src/harness/execution/step_handlers/variables.py:34
↓ 3 callersMethodexecute_llm_step
Execute a single LLM step with tool calling loop. Prepares messages and tools, then delegates to multi_step_tool_call_loop for the ac
src/harness/execution/llm_executor.py:194
↓ 3 callersMethodexecute_switch_step
Execute switch/case style multi-way branching. Selects and executes one branch based on a variable's value. Falls back to default bra
src/harness/execution/step_handlers/control.py:376
↓ 3 callersFunctionexpand
(value: str)
src/harness/utils/step_utils.py:22
↓ 3 callersFunctionextract_answer
Extract the answer from [ANSWER]...[/ANSWER] tags in the response.
src/benchmarks/chembench/evaluate.py:29
↓ 3 callersFunctionextract_code_blocks_with_lang
(text: str)
src/harness/parsing/inline_tools.py:33
↓ 3 callersFunctiongetProjectRoot
()
yaml-flow-editor/vite.config.ts:65
↓ 3 callersMethodget_session_id
Return the current MCP Streamable HTTP session id, if known. This is the value that must be persisted in checkpoints to allow resumi
src/mcp_client/client.py:129
↓ 3 callersMethodget_summary
Human-readable summary for the orchestrator LLM.
src/harness/agentic_loop/task_board.py:119
↓ 3 callersMethodget_task
(self, task_id: str)
src/harness/agentic_loop/task_board.py:104
↓ 3 callersMethodget_usage_summary
Return a summary of token usage.
src/harness/agentic_loop/session.py:64
↓ 3 callersFunctiongraphToYaml
(nodes: FlowNode[], edges: FlowEdge[])
yaml-flow-editor/src/converters/graph-to-yaml.ts:395
↓ 3 callersFunctionhandle_request
(message_text, channel, thread_ts, user_id, error_callback=None)
scripts/slack_listener.py:283
↓ 3 callersFunctionload_env_file
(env_path: Path)
scripts/discord_listener.py:50
↓ 3 callersFunctionload_env_file
(env_path: Path)
scripts/slack_listener.py:53
↓ 3 callersMethodload_initial
(self)
scripts/dashboard.py:445
↓ 3 callersMethodminimal_env
(self)
src/sandbox_vm/tools/session_management.py:136
↓ 3 callersFunctionnormalizeObject
(obj: T | null | undefined)
yaml-flow-editor/src/utils/workflowStepConverters.ts:44
↓ 3 callersMethodprint_messages
Print and log messages. Thread-safe for parallel execution.
src/mcp_client/logger.py:938
↓ 3 callersFunctionreadBody
(req: IncomingMessage, limitBytes = 2 * 1024 * 1024)
yaml-flow-editor/vite.config.ts:31
↓ 3 callersFunctionreadRunMeta
(projectRoot: string, runId: string)
yaml-flow-editor/vite.config.ts:102
↓ 3 callersFunctionresolve_system_prompt
Resolve a system prompt with template expansion.
src/harness/prompts/yaml_prompts.py:29
↓ 3 callersMethodrestore_workspace
Restore workspace from backup during resume using rsync. Falls back to shutil.copytree if rsync is not available.
src/harness/checkpoints/checkpoint_manager.py:192
↓ 3 callersFunctionrun_loop
Start the interactive agentic loop against a running sandbox VM.
src/harness/agentic_loop/run_loop.py:51
↓ 3 callersFunctionsafe_json_dumps
Serialize data to JSON with support for complex types. Handles serialization of: - Dataclasses (converted to dict) - Sets and tuples (con
src/harness/types/serialization.py:46
↓ 3 callersFunctionsandbox_query_dir
Return the sandbox-side output directory for a specific query. Mirrors the host-side layout so paths are easy to find:: outputs/{experim
src/benchmarks/writing_bench/instance.py:25
↓ 3 callersFunctionshorten_values
(msg: Union[List[Any], Dict[str, Any]], max_value_len=80)
src/utils.py:18
↓ 3 callersFunctionsplit_step_value
(step_data: Dict[str, Any], key: str)
src/harness/utils/step_utils.py:7
↓ 3 callersMethodsubscribe
(self, sid: str, path: Optional[str])
scripts/dashboard.py:563
↓ 3 callersMethodto_dict
(self)
src/harness/agentic_loop/task_board.py:173
↓ 3 callersFunctiontools_to_openai_format
Merge orchestrator and MCP tools into OpenAI function-calling format.
src/harness/agentic_loop/tools.py:143
↓ 3 callersMethodtouch
(self)
src/sandbox_vm/tools/session_management.py:133
↓ 3 callersFunctiontrigger_shutdown
Stop the server and exit when there are no more clients.
scripts/dashboard.py:621
↓ 3 callersFunctionupdateCall
(index: number, field: keyof GatherCall, value: unknown)
yaml-flow-editor/src/components/NodeConfigPanel.tsx:215
↓ 3 callersFunctionupdateModuleWorkflowYaml
( originalContent: string, newSteps: FlowNodeData[], )
yaml-flow-editor/src/utils/moduleYaml.ts:9
↓ 3 callersMethodvalidate_yaml_syntax
Validate YAML syntax
src/harness/parsing/yaml_parser.py:271
↓ 2 callersFunction_append_jsonl
(path: str, event: Dict[str, Any])
src/mcp_client/client.py:15
↓ 2 callersFunction_backup_large_file
Copy a file to *backup_dir* preserving its relative path. Returns the absolute path inside backup_dir on success, None on failure.
src/harness/checkpoints/workspace_manifest.py:75
↓ 2 callersMethod_call_llm
Single LLM call with system prompt, no tools.
src/harness/agentic_loop/plan_generator.py:166
↓ 2 callersFunction_capture_vision
Capture a screenshot and optionally annotate it with Set-of-Marks. Returns a dict with zero or more of: screenshot_filename: vpath to the r
src/sandbox_vm/tools/app_control/browser.py:730
↓ 2 callersFunction_clamp_timeout
(t: int, low: int = 1, high: int = 300)
src/sandbox_vm/tools/code/python.py:33
↓ 2 callersMethod_configure_tracing
Configure tracing state for the LLMExecutor. Args: trace_writer: TraceWriter for recording events replay_events: List
src/harness/agent.py:156
↓ 2 callersFunction_copy_submodules_for_workflow
Copy all submodule YAMLs referenced (recursively) by the workflow. Submodules are copied under a `submodules/` directory inside the reproduci
src/harness/utils/reproducibility.py:246
↓ 2 callersFunction_download_file
Download a file from a URL. Returns True on success.
src/benchmarks/writing_bench/run.py:43
↓ 2 callersFunction_ensure_loaded
Lazy-load the BM25 index and corpus lookup on first use.
src/sandbox_vm/tools/browsecomp/tools.py:33
↓ 2 callersMethod_env_bool
(key: str, *, default: bool)
src/mcp_client/logger.py:194
↓ 2 callersMethod_exec_shell_run
Execute a shell command via docker exec.
src/benchmarks/swe_bench_verified/docker_exec_client.py:174
↓ 2 callersMethod_expand_template
Expand template variables
verifier/AgentVerifier/yaml_parser.py:252
↓ 2 callersFunction_extract
(page, state: _SessionState, *, limit: int)
src/sandbox_vm/tools/app_control/browser.py:457
↓ 2 callersFunction_extract_boxed
r"""Extract all \boxed{...} contents, handling nested braces.
src/benchmarks/aime/evaluate.py:40
↓ 2 callersMethod_extract_yaml
Extract YAML from a ```yaml fenced code block, or use raw text.
src/harness/agentic_loop/plan_generator.py:181
↓ 2 callersFunction_format_function_call
(name: str, args: tuple, kwargs: dict)
src/sandbox_vm/mcp/server.py:20
↓ 2 callersFunction_friendly_step_name
(raw_name: str, module_path: str = "")
demo/ai_scientists/web/app.py:228
↓ 2 callersFunction_get_min_indentation
Get the minimum indentation from a list of lines (the baseline). Args: lines: List of code lines Returns: The minimum i
src/sandbox_vm/tools/code/patch.py:218
↓ 2 callersFunction_get_model_completions
()
src/harness/ephemeral/repl.py:50
↓ 2 callersMethod_handle_id
(self, handle_uri: str)
src/sandbox_vm/tools/session_management.py:212
↓ 2 callersFunction_infer_param_type
Infer Lean BaseType from a JSON parameter value.
verifier/AgentVerifier/WorkflowToLean.py:354
↓ 2 callersFunction_infer_set_variable_type
Infer set_variable output type; keep passthrough type when possible.
verifier/AgentVerifier/WorkflowToLean.py:369
↓ 2 callersMethod_inner_loop_with_tools
Run the inner loop with a custom tool set. Temporarily swaps tool lists, runs _inner_loop, then restores. Used by plan mode phases to
src/harness/agentic_loop/loop.py:883
↓ 2 callersMethod_invoke_submodule_tool
Invoke a submodule as a tool call. Looks up the submodule specification, resolves parameters from context and tool arguments, then ex
src/harness/execution/llm_executor.py:259
↓ 2 callersMethod_is_submodule_submit_result
Check if a submodule result contains the submit completion signal. When ``require_submodule_submit`` is enabled, the submit_changes s
src/harness/execution/shell_utils.py:86
↓ 2 callersFunction_is_text_file
Heuristic check: read first *sample_size* bytes and look for null bytes.
src/harness/checkpoints/workspace_manifest.py:38
↓ 2 callersFunction_iter_workflow_files
()
scripts/discord_listener.py:39
↓ 2 callersFunction_iter_workflow_files
()
scripts/slack_listener.py:42
↓ 2 callersMethod_kv_preview_table
Compact preview for big dict payloads.
src/mcp_client/logger.py:238
↓ 2 callersMethod_leave_room
(self, sid: str, path_str: str)
scripts/dashboard.py:592
↓ 2 callersFunction_load_env_file
Load KEY=VALUE pairs from an env file into os.environ.
verifier/AgentVerifier/WorkflowToLean.py:307
↓ 2 callersMethod_make_args
(self, **overrides)
tests/tool_tests/test_inline_tool_calls.py:195
↓ 2 callersMethod_make_executor
(self)
tests/tool_tests/test_inline_tool_calls.py:179
↓ 2 callersMethod_maybe_apply_side_effects_in_replay
( self, name: str, arguments: Dict[str, Any] )
src/mcp_client/client.py:166
↓ 2 callersMethod_next
(self)
src/harness/checkpoints/durable_execution.py:55
↓ 2 callersFunction_normalize_param_list
(items: Any, source: str)
src/harness/submodules/loader.py:68
↓ 2 callersMethod_parse_basic_step
Parse step (with conversation history)
src/harness/parsing/yaml_parser.py:174
↓ 2 callersFunction_parse_events
(event_log_path: str)
demo/ai_scientists/web/app.py:255
↓ 2 callersMethod_parse_for_each_step
Parse loop step
src/harness/parsing/yaml_parser.py:192
↓ 2 callersMethod_parse_if_step
Parse conditional step
src/harness/parsing/yaml_parser.py:204
↓ 2 callersMethod_parse_increment_step
Parse increment step
src/harness/parsing/yaml_parser.py:239
↓ 2 callersMethod_parse_switch_step
Parse switch statement
src/harness/parsing/yaml_parser.py:222
↓ 2 callersMethod_parse_while_step
Parse while loop
src/harness/parsing/yaml_parser.py:213
↓ 2 callersMethod_persist
(self)
src/harness/agentic_loop/task_board.py:183
↓ 2 callersFunction_persistent_mcp_url
()
src/harness/ephemeral/repl.py:136
↓ 2 callersMethod_read_full
(self)
scripts/dashboard.py:433
← previousnext →201–300 of 1,694, ranked by callers