| 1 | # skills.py DOX |
| 2 | |
| 3 | ## Purpose |
| 4 | |
| 5 | - Own the `skills.py` helper module. |
| 6 | - This module discovers, parses, filters, and resolves Agent Zero skills. |
| 7 | - Keep this file-level DOX profile synchronized with `skills.py` because this directory is intentionally flat. |
| 8 | |
| 9 | ## Ownership |
| 10 | |
| 11 | - `skills.py` owns the runtime implementation. |
| 12 | - `skills.py.dox.md` owns durable notes about responsibilities, contracts, side effects, and verification for that implementation. |
| 13 | - Classes: |
| 14 | - `ActiveSkillEntry` (`TypedDict`) |
| 15 | - `CatalogSkill` (`TypedDict`) |
| 16 | - `Skill` (no explicit base class) |
| 17 | - Top-level functions: |
| 18 | - `get_skills_base_dir() -> Path` |
| 19 | - `get_skill_roots(agent: Agent | None=...) -> List[str]` |
| 20 | - `_is_hidden_path(path: Path) -> bool` |
| 21 | - `discover_skill_md_files(root: Path) -> List[Path]`: Recursively discover SKILL.md files under a root directory. |
| 22 | - `_coerce_list(value: Any) -> List[str]` |
| 23 | - `_normalize_name(name: str) -> str` |
| 24 | - `_read_text(path: Path) -> str` |
| 25 | - `split_frontmatter(markdown: str) -> Tuple[Dict[str, Any], str, List[str]]`: Splits a SKILL.md into (frontmatter_dict, body_text, errors). |
| 26 | - `_parse_frontmatter_fallback(frontmatter_text: str) -> Dict[str, Any]` |
| 27 | - `parse_frontmatter(frontmatter_text: str) -> Tuple[Dict[str, Any], List[str]]`: Parse YAML frontmatter with PyYAML when available, |
| 28 | - `skill_from_markdown(skill_md_path: Path, include_content: bool=..., validate: bool=...) -> Optional[Skill]` |
| 29 | - `list_skills(agent: Agent | None=..., include_content: bool=..., include_hidden: bool=...) -> List[Skill]`: List skills, optionally filtered by agent scope. |
| 30 | - `delete_skill(skill_path: str) -> None`: Delete a skill directory. |
| 31 | - `find_skill(skill_name: str, agent: Agent | None=..., include_content: bool=..., include_hidden: bool=..., validate: bool=...) -> Optional[Skill]` |
| 32 | - `load_skill_for_agent(skill_name: str, agent: Agent | None=...) -> str`: Load skill and format it as a complete string for agent context. |
| 33 | - `skill_instruction_name(message: Any) -> str` |
| 34 | - `_get_skill_files(skill_dir: Path) -> str`: Get file tree for skill directory. |
| 35 | - `search_skills(query: str, limit: int=..., agent: Agent | None=..., include_hidden: bool=...) -> List[Skill]` |
| 36 | - `validate_skill(skill: Skill) -> List[str]` |
| 37 | - `validate_skill_md(skill_md_path: Path) -> List[str]` |
| 38 | - `_normalize_max_active_skills(value: Any) -> int` |
| 39 | - `get_max_active_skills(agent: Agent | None=..., project_name: str | None=...) -> int` |
| 40 | - `normalize_skills_config(config: dict[str, Any] | None) -> dict[str, Any]` |
| 41 | - `normalize_visibility_policy(raw: Any) -> dict[str, Any]` |
| 42 | - `get_visibility_policy(agent: Agent | None) -> dict[str, Any]` |
| 43 | - `is_skill_allowed(policy, skill_or_entry) -> bool` |
| 44 | - `ensure_skill_visible(agent, entry) -> None` |
| 45 | - `normalize_active_skills(raw: Any, limit: int | None=...) -> list[ActiveSkillEntry]` |
| 46 | - `normalize_hidden_skills(raw: Any) -> list[ActiveSkillEntry]` |
| 47 | - `normalize_skill_entries(raw: Any, limit: int | None=...) -> list[ActiveSkillEntry]` |
| 48 | - `list_skill_catalog(project_name: str=..., agent: Agent | None=...) -> list[CatalogSkill]` |
| 49 | - `get_scope_active_skills(agent: Agent | None) -> list[ActiveSkillEntry]` |
| 50 | - `get_scope_hidden_skills(agent: Agent | None) -> list[ActiveSkillEntry]` |
| 51 | - `get_chat_active_skills(context: Any | None) -> list[ActiveSkillEntry]` |
| 52 | - `get_chat_disabled_skills(context: Any | None) -> list[ActiveSkillEntry]` |
| 53 | - `get_loaded_skill_names(agent: Agent | None) -> list[str]` |
| 54 | - `set_loaded_skill_names(agent: Agent | None, skill_names: Any) -> list[str]` |
| 55 | - `add_loaded_skill_name(agent: Agent | None, skill_name: str, limit: int | None=...) -> list[str]` |
| 56 | - Notable constants/configuration names: `MAX_ACTIVE_SKILLS`, `ACTIVE_SKILLS_PLUGIN_NAME`, `AGENT_DATA_NAME_LOADED_SKILLS`, `CONTEXT_DATA_NAME_LOADED_SKILLS`, `CONTEXT_DATA_NAME_CHAT_ACTIVE_SKILLS`, `CONTEXT_DATA_NAME_CHAT_DISABLED_SKILLS`, `CONTEXT_DATA_NAME_CHAT_VISIBLE_SKILLS`, `_NAME_RE`. |
| 57 | |
| 58 | ## Runtime Contracts |
| 59 | |
| 60 | - Helper modules own reusable framework APIs and must preserve public callers unless all callers, tests, and docs are updated together. |
| 61 | - Update this file whenever public functions, classes, persistence behavior, path/security assumptions, side effects, or cross-module contracts change. |
| 62 | - Loaded skill names are chat-wide context data under `CONTEXT_DATA_NAME_LOADED_SKILLS`; legacy agent-local `loaded_skills` lists are migrated into context data and cleared when read. |
| 63 | - Loaded skill bodies live in chat history; hiding a skill changes catalog visibility but does not remove the loaded-skill ledger. |
| 64 | - `_skills.visibility_policy` is profile-aware and uses explicit future-skill |
| 65 | allow/block defaults. It filters discovery, search, new loading, chat |
| 66 | activation, and active-skill resolution without removing instructions already |
| 67 | stored in chat history. |
| 68 | - Visibility policy IDs match both canonical skill paths and their directory |
| 69 | names, and bulk discovery resolves the effective policy only once. |
| 70 | - Legacy `hidden_skills` remains default-allow with blocked exceptions; a chat |
| 71 | visibility override cannot bypass profile visibility policy. |
| 72 | - `build_active_skills_prompt()` returns empty because selected skills are loaded through history, not prompt protocol. |
| 73 | - `search_skills()` normalizes query words, scores normal terms against skill names, and scores only long terms against tags/triggers; descriptions match only full query phrases so generic prose does not produce irrelevant suggestions. |
| 74 | - `find_skill(validate=False)` lets validation tooling resolve a skill with incomplete metadata while preserving runtime validation by default. |
| 75 | - Slash command discovery is delegated to the built-in `_commands` helper through a local import, preserving its project/global/bundled/plugin precedence and avoiding its `split_frontmatter` import cycle. Picker-hidden commands remain hidden from Skills; reading a command returns its definition without executing it. |
| 76 | - Invalid `SKILL.md` frontmatter emits a once-per-path scan warning with the skipped skill path/name and a line number when the parser can identify one directly. |
| 77 | - Observed side-effect areas: filesystem reads, filesystem deletion, plugin state, settings/state persistence, context data, secret handling. |
| 78 | - Imported dependency areas include: `__future__`, `dataclasses`, `helpers`, `os`, `pathlib`, `re`, `typing`. |
| 79 | |
| 80 | ## Key Concepts |
| 81 | |
| 82 | - Important called helpers/classes observed in the source: `dataclass`, `re.compile`, `field`, `Path`, `root.rglob`, `results.sort`, `re.sub`, `path.read_text`, `text.splitlines`, `join.strip`, `parse_frontmatter`, `frontmatter_text.splitlines`, `_parse_frontmatter_fallback`, `split_frontmatter`, `str.strip`, `_coerce_list`, `Skill`, `get_skill_roots`, `_filter_hidden_skills`, `files.get_abs_path`. |
| 83 | - Keep request/response, tool, or helper semantics documented here at the same time as source changes. |
| 84 | |
| 85 | ## Work Guidance |
| 86 | |
| 87 | - Preserve public helper APIs used by core code and plugins unless every caller is updated. |
| 88 | - Keep path, auth, secret, persistence, network, and subprocess behavior explicit and bounded. |
| 89 | - Prefer adding cohesive helper functions here only when behavior is reused across modules. |
| 90 | |
| 91 | ## Verification |
| 92 | |
| 93 | - Run targeted tests for changed helper behavior; run security regressions for auth, filesystem, WebSocket, tunnel, upload, or secret-handling helpers. |
| 94 | - Related tests observed by source search: |
| 95 | - `tests/test_a0_connector_prompt_gating.py` |
| 96 | - `tests/test_browser_agent_regressions.py` |
| 97 | - `tests/test_document_query_plugin.py` |
| 98 | - `tests/test_fasta2a_client.py` |
| 99 | - `tests/test_office_canvas_setup.py` |
| 100 | - `tests/test_office_document_store.py` |
| 101 | - `tests/test_skills_runtime.py` |
| 102 | - `tests/test_time_travel.py` |
| 103 | |
| 104 | ## Child DOX Index |
| 105 | |
| 106 | No child DOX files. |