Functions — github.com/Ericbai06/SkillDAG

hub / github.com/Ericbai06/SkillDAG / functions

Functions270 in github.com/Ericbai06/SkillDAG

⨍ Functions 270◇ Types & classes 11

↓ 36 callersMethod_reload

(self)

tests/test_graph.py:51

↓ 28 callersMethodkey

(self)

src/skilldag/graph.py:139

↓ 20 callersMethodsave

(self)

src/skilldag/graph.py:260

↓ 19 callersMethod_create_skill

(self, skill_id: str, name: str, description: str)

tests/test_graph.py:43

↓ 19 callersFunction_emit_json

(payload: dict, exit_code: int = 0)

src/skilldag/cli.py:21

↓ 15 callersFunction_graph_context

(args)

src/skilldag/cli.py:148

↓ 14 callersMethod_normalize_edge

(self, source: str, target: str, edge_type: str)

src/skilldag/graph.py:315

↓ 14 callersMethodsearch

Vector top-K matches plus a separate depth-bounded neighbor list. Three sections, no score mixing: - ``matches``: top ``top_k`` skil

src/skilldag/graph.py:416

↓ 13 callersMethodedit_edge

Commit an add / remove / retype immediately.

src/skilldag/graph.py:1072

↓ 11 callersFunctionfail

(message: str)

scripts/validate_graph_inputs.py:13

↓ 11 callersMethodload

( cls, graph_path: str | Path = DEFAULT_GRAPH_PATH, skills_dir: str | Path = DEFAULT_S

src/skilldag/graph.py:150

↓ 9 callersMethod_ensure_skill_exists

(self, skill_id: str)

src/skilldag/graph.py:400

↓ 7 callersFunction_normalize

(vec: list[float])

src/skilldag/initialize.py:335

↓ 6 callersMethod_edge_lookup

(self)

src/skilldag/graph.py:326

↓ 6 callersFunctionread_json

(path: Path)

benchmarks/skill_use_metrics.py:109

↓ 5 callersFunction_embed_batch

Call embedding API in batches; return list aligned with input order. Per-batch retries on Timeout / 429 / 5xx with exponential backoff.

src/skilldag/initialize.py:278

↓ 5 callersMethod_patch_embedding_backend

Install a deterministic embedding backend for offline search tests.

tests/test_graph.py:55

↓ 5 callersMethod_validate_reason

(reason: str)

src/skilldag/graph.py:960

↓ 5 callersFunctionload_json

(path: Path)

scripts/replay_queries.py:42

↓ 5 callersMethodpropose_edge

Dry-run add. Returns would/related; never mutates.

src/skilldag/graph.py:966

↓ 5 callersMethodpropose_retype_edge

Dry-run retype. Returns would/related; never mutates.

src/skilldag/graph.py:1008

↓ 4 callersMethod_append_history

( self, action: str, reason: str, task_id: str, *, edge: dict[

src/skilldag/graph.py:1045

↓ 4 callersMethod_assert_acyclic

(self, source: str, target: str, edge_type: str)

src/skilldag/graph.py:360

↓ 4 callersMethod_assert_no_contradiction

( self, source: str, target: str, edge_type: str, *, ignore_ed

src/skilldag/graph.py:376

↓ 4 callersFunction_shorten

(text: str, limit: int)

analysis/backfill_trial_logs.py:17

↓ 4 callersFunctionas_float

(value: Any)

analysis/score_skillsbench_gos.py:66

↓ 4 callersFunctionquantiles

(values: list[float])

analysis/analyze_skilldag_run.py:55

↓ 4 callersFunctionutc_now_iso

()

src/skilldag/graph.py:80

↓ 3 callersMethod_gather_related

(self, src: str, tgt: str)

src/skilldag/graph.py:907

↓ 3 callersFunction_http_post_json

POST JSON via stdlib urllib. Returns (status, body_text). Raises ``_HTTPTimeout`` on socket timeout, ``RuntimeError`` on any other transport-

src/skilldag/initialize.py:65

↓ 3 callersFunction_require_env

(name: str)

src/skilldag/initialize.py:233

↓ 3 callersFunction_task_name_from_trial_dir

(path: Path)

benchmarks/skill_use_metrics.py:117

↓ 3 callersMethodcheck_set

(self, skill_ids: list[str])

src/skilldag/graph.py:850

↓ 3 callersMethodexpand_set

(self, skill_ids: list[str])

src/skilldag/graph.py:829

↓ 3 callersFunctionextract_skill_ids_from_text

(text: str)

benchmarks/skill_use_metrics.py:121

↓ 3 callersFunctioninitialize_edges

Build cold-start edges for the given nodes dict (skill_id -> node metadata). Args: nodes: {skill_id: {name, description, path, status, ta

src/skilldag/initialize.py:677

↓ 3 callersFunctionis_trial_dir

(path: Path)

analysis/score_skillsbench_gos.py:152

↓ 3 callersFunctionload_json

(path: Path)

analysis/analyze_skilldag_run.py:30

↓ 3 callersFunctionload_json

(path: Path)

analysis/score_skillsbench_gos.py:57

↓ 3 callersMethodpropose_remove_edge

Dry-run remove. Returns would/related; never mutates.

src/skilldag/graph.py:989

↓ 3 callersFunctionretrieval_score

Run a single query against a graph and score it.

scripts/replay_queries.py:182

↓ 3 callersFunctionsave_json

(path: Path, data: Any)

scripts/replay_queries.py:47

↓ 3 callersFunctionwrite_alfworld_skill_use_metrics

( result_dirs: Iterable[Path], *, output: Path | None = None, alfworld_data: Path | None = Non

benchmarks/skill_use_metrics.py:466

↓ 2 callersFunction_classify_anchor_batch

Classify anchor vs every candidate in one LLM call. Returns a list of edge dicts {source, target, type, reason} — only non-"none" results. In

src/skilldag/initialize.py:523

↓ 2 callersMethod_collect_conflicts

(self, match_ids: list[str])

src/skilldag/graph.py:708

↓ 2 callersMethod_collect_neighbors

( self, match_ids: list[str], depth: int )

src/skilldag/graph.py:670

↓ 2 callersMethod_discover_nodes

(skills_dir: Path)

src/skilldag/graph.py:212

↓ 2 callersMethod_edge_polarity

``conflicts_with`` marks "do not co-select"; every other type asserts a positive relation. The two polarities are mutually exclusive f

src/skilldag/graph.py:370

↓ 2 callersMethod_format_match_entry

(self, skill_id: str, *, score: float)

src/skilldag/graph.py:746

↓ 2 callersMethod_load_or_build_node_embeddings

Return ``{skill_id: embedding}`` for every current node. Cache file ``<graph_path>.embeddings.json`` keyed by skill_id stores ``{text

src/skilldag/graph.py:543

↓ 2 callersFunction_read_body_preview

Return first max_chars of SKILL.md body (after frontmatter, skipping headings).

src/skilldag/initialize.py:247

↓ 2 callersFunction_read_json

(path: Path)

analysis/backfill_trial_logs.py:24

↓ 2 callersFunction_strip_json_fence

(text: str)

src/skilldag/initialize.py:432

↓ 2 callersMethodas_dict

(self)

src/skilldag/graph.py:132

↓ 2 callersMethodas_dict

(self)

benchmarks/skill_use_metrics.py:92

↓ 2 callersFunctionbase_task

(trial_dir: Path)

analysis/analyze_skilldag_run.py:81

↓ 2 callersFunctionbuild_payload

(rows: list[SkillUseRow], *, gold_source: str)

benchmarks/skill_use_metrics.py:419

↓ 2 callersFunctioncli_reference

()

benchmarks/shared/skilldag_prompt.py:194

↓ 2 callersFunctioncompute_alfworld_skill_use_row

( result: dict[str, Any], *, source: str, alfworld_data: Path, include_navigation: bool =

benchmarks/skill_use_metrics.py:225

↓ 2 callersFunctiondefault_alfworld_data

()

benchmarks/skill_use_metrics.py:113

↓ 2 callersFunctionduration_seconds

(block: dict[str, Any] | None)

analysis/analyze_skilldag_run.py:45

↓ 2 callersFunctionextract_skill_ids

(texts: list[str], pattern: str)

analysis/analyze_skilldag_run.py:182

↓ 2 callersFunctionextract_used_skills_from_result

(result: dict[str, Any])

benchmarks/skill_use_metrics.py:131

↓ 2 callersFunctionfailure_reflection_protocol

()

benchmarks/shared/skilldag_prompt.py:198

↓ 2 callersFunctionfew_shot

()

benchmarks/shared/skilldag_prompt.py:206

↓ 2 callersFunctionformat_counts

(counts: dict[str, int])

analysis/score_skillsbench_gos.py:300

↓ 2 callersMethodget_alternatives

(self, skill_id: str)

src/skilldag/graph.py:813

↓ 2 callersMethodget_conflicts

(self, skill_id: str)

src/skilldag/graph.py:802

↓ 2 callersMethodget_dependencies

(self, skill_id: str, transitive: bool = False)

src/skilldag/graph.py:770

↓ 2 callersFunctionhardlink_or_copy_file

(source: Path, destination: Path)

benchmarks/skillsbench/graphskills_benchmark.py:30

↓ 2 callersFunctionis_openrouter_chat_model

(model_name: str, api_base: str | None = None)

benchmarks/alfworld/skilldag_runtime.py:115

↓ 2 callersFunctionjson_objects

(text: str)

scripts/replay_queries.py:54

↓ 2 callersFunctionload_json

(path: Path)

scripts/validate_graph_inputs.py:22

↓ 2 callersFunctionmain

()

src/skilldag/cli.py:472

↓ 2 callersFunctionmean_rank

(rows: list[dict[str, Any]])

scripts/replay_queries.py:275

↓ 2 callersFunctionmutation_rules

()

benchmarks/shared/skilldag_prompt.py:210

↓ 2 callersFunctionparse_dt

(value: str | None)

analysis/analyze_skilldag_run.py:39

↓ 2 callersFunctionpct

(p: float)

analysis/analyze_skilldag_run.py:60

↓ 2 callersFunctionsearch_protocol

()

benchmarks/shared/skilldag_prompt.py:214

↓ 2 callersFunctionsuccess_path_discovery

()

benchmarks/shared/skilldag_prompt.py:218

↓ 2 callersFunctiontrial_dirs

(job_dir: Path)

analysis/analyze_skilldag_run.py:77

↓ 2 callersFunctionwarn

(message: str)

scripts/validate_graph_inputs.py:18

↓ 2 callersFunctionwhen_to_mutate

()

benchmarks/shared/skilldag_prompt.py:202

↓ 2 callersFunctionwrite_payload

(payload: dict[str, Any], output: Path | None)

benchmarks/skill_use_metrics.py:458

↓ 1 callersMethod__init__

(self, graph_path: Path, skills_dir: Path, data: dict[str, Any])

src/skilldag/graph.py:144

↓ 1 callersFunction_alfworld_system_prompt

()

benchmarks/alfworld/skilldag_runtime.py:347

↓ 1 callersMethod_backbone_reachable

True if ``dst`` is reachable from ``src`` along directed backbone edges (``depends_on`` / ``specializes``) only.

src/skilldag/graph.py:341

↓ 1 callersFunction_build_graph_context

Return a compact graph-context block for skill_id, or '' if no edges / graph absent.

src/skilldag/cli.py:30

↓ 1 callersFunction_cos_matrix

Symmetric cos-sim matrix; diagonal left as 0 (we skip self anyway).

src/skilldag/initialize.py:340

↓ 1 callersMethod_embed_queries_batch

Embed N queries in one API call.

src/skilldag/graph.py:622

↓ 1 callersMethod_embed_query

(self, query: str)

src/skilldag/graph.py:613

↓ 1 callersMethod_embeddings_cache_path

(self)

src/skilldag/graph.py:528

↓ 1 callersFunction_ensure_skilldag_wrapper

Expose `skilldag` as a normal command on PATH for the shell tool.

benchmarks/alfworld/skilldag_runtime.py:413

↓ 1 callersFunction_extract_exception

(result: dict[str, Any])

analysis/backfill_trial_logs.py:41

↓ 1 callersFunction_extract_first_json_object

(response: str)

benchmarks/alfworld/skilldag_runtime.py:259

↓ 1 callersFunction_extract_reward

(result: dict[str, Any])

analysis/backfill_trial_logs.py:33

↓ 1 callersFunction_extract_skill_id

Best-effort: if the CLI command references a skill id (as a positional arg to ``skilldag show`` or ``skilldag graph <sub> <id>``), return that

benchmarks/alfworld/skilldag_runtime.py:234

↓ 1 callersMethod_find_neighbors

(self, skill_id: str, edge_type: str | None = None)

src/skilldag/graph.py:404

↓ 1 callersFunction_format_trial_log

(trial_dir: Path, max_field_chars: int)

analysis/backfill_trial_logs.py:50

↓ 1 callersFunction_gold_dir_for_trial

(trial_dir: Path, result: dict[str, Any] | None, gold_tasks_dir: Path)

benchmarks/skill_use_metrics.py:311

next →1–100 of 270, ranked by callers