ty checking
This commit is contained in:
John Lancaster
@@ -30,15 +30,9 @@ class SkillFilesBundle:
|
||||
@classmethod
|
||||
def from_paths(cls, slug: str, paths: set[MarkdownDocument]) -> Self:
|
||||
skill = next(iter(p for p in paths if p.relpath.name == "SKILL.md"))
|
||||
groups = {
|
||||
k: tuple(v)
|
||||
for k, v in groupby(
|
||||
paths,
|
||||
key=lambda p: fnmatch(p.relpath.as_posix(), f"skills/{slug}/references/*.md"),
|
||||
)
|
||||
}
|
||||
references = groups.get(True, ())
|
||||
other = tuple(p for p in groups.get(False, ()) if p != skill)
|
||||
sorted_paths = tuple(sorted(paths, key=lambda p: p.relpath.as_posix()))
|
||||
references = tuple(p for p in sorted_paths if fnmatch(p.relpath.as_posix(), f"skills/{slug}/references/*.md"))
|
||||
other = tuple(p for p in sorted_paths if p not in references and p != skill)
|
||||
return cls(
|
||||
slug=slug,
|
||||
skill=skill,
|
||||
|
||||
+279
-187
@@ -1,6 +1,14 @@
|
||||
from importlib.resources import files
|
||||
from pathlib import PurePosixPath
|
||||
|
||||
from pydantic import ValidationError
|
||||
|
||||
from ..skills.document_loader import _discover_top_level_references
|
||||
from ..skills.document_loader import _parse_frontmatter
|
||||
from ..skills.document_loader import _reference_id_from_filename
|
||||
from ..skills.document_loader import _validate_prompt_frontmatter
|
||||
from ..skills.document_loader import _validate_skill_frontmatter
|
||||
from ..skills.document_loader import _walk_markdown
|
||||
from .contracts import DocsRegistry
|
||||
from .contracts import PromptRecord
|
||||
from .contracts import PromptSummaryRecord
|
||||
@@ -11,49 +19,47 @@ from .issues import DocsRegistryValidationError
|
||||
from .issues import RegistryIssue
|
||||
|
||||
|
||||
def load_docs_registry(
|
||||
def _ensure_no_cycles(
|
||||
skills_by_id: dict[str, SkillRecord],
|
||||
) -> list[tuple[str, str]]:
|
||||
state: dict[str, int] = {}
|
||||
stack: list[str] = []
|
||||
cycles: set[tuple[str, str]] = set()
|
||||
|
||||
def visit(skill_id: str) -> None:
|
||||
state[skill_id] = 1
|
||||
stack.append(skill_id)
|
||||
|
||||
for dependency in skills_by_id[skill_id].depends_on:
|
||||
if dependency not in skills_by_id:
|
||||
continue
|
||||
dep_state = state.get(dependency, 0)
|
||||
if dep_state == 0:
|
||||
visit(dependency)
|
||||
continue
|
||||
if dep_state == 1:
|
||||
cycle_start = stack[stack.index(dependency)]
|
||||
cycle_path = stack[stack.index(dependency) :] + [dependency]
|
||||
cycles.add((cycle_start, " -> ".join(cycle_path)))
|
||||
|
||||
stack.pop()
|
||||
state[skill_id] = 2
|
||||
|
||||
for skill_id in sorted(skills_by_id):
|
||||
if state.get(skill_id, 0) == 0:
|
||||
visit(skill_id)
|
||||
|
||||
return sorted(cycles)
|
||||
|
||||
|
||||
def _load_skills(
|
||||
*,
|
||||
package_anchor: str,
|
||||
docs_root: str = "docs",
|
||||
) -> DocsRegistry:
|
||||
docs_dir = files(package_anchor).joinpath(docs_root)
|
||||
issues: list[RegistryIssue] = []
|
||||
|
||||
if not docs_dir.is_dir():
|
||||
raise DocsRegistryValidationError(
|
||||
[
|
||||
RegistryIssue(
|
||||
code="missing_docs_root",
|
||||
message="docs root directory does not exist",
|
||||
skill_id=None,
|
||||
path=docs_root,
|
||||
hint="configure docs_root to a valid packaged docs path",
|
||||
)
|
||||
]
|
||||
)
|
||||
|
||||
docs_markdown_by_path: dict[str, str] = {}
|
||||
for relpath, doc_file in _walk_markdown(docs_dir):
|
||||
docs_markdown_by_path[relpath] = doc_file.read_text(encoding="utf-8")
|
||||
|
||||
skills_root = docs_dir.joinpath("skills")
|
||||
if not skills_root.is_dir():
|
||||
raise DocsRegistryValidationError(
|
||||
[
|
||||
RegistryIssue(
|
||||
code="missing_skills_root",
|
||||
message="skills directory does not exist under docs root",
|
||||
skill_id=None,
|
||||
path=f"{docs_root}/skills",
|
||||
hint="ensure docs/skills is included in packaged docs",
|
||||
)
|
||||
]
|
||||
)
|
||||
|
||||
docs_markdown_by_path: dict[str, str],
|
||||
skills_root,
|
||||
) -> tuple[dict[str, SkillRecord], list[SkillSummaryRecord], list[RegistryIssue]]:
|
||||
skills_by_id: dict[str, SkillRecord] = {}
|
||||
summaries: list[SkillSummaryRecord] = []
|
||||
prompts_by_id: dict[str, PromptRecord] = {}
|
||||
prompt_summaries: list[PromptSummaryRecord] = []
|
||||
issues: list[RegistryIssue] = []
|
||||
|
||||
for skill_dir in sorted(skills_root.iterdir(), key=lambda item: item.name):
|
||||
if not skill_dir.is_dir():
|
||||
@@ -164,153 +170,173 @@ def load_docs_registry(
|
||||
)
|
||||
)
|
||||
|
||||
prompts_root = docs_dir.joinpath("prompts")
|
||||
return skills_by_id, summaries, issues
|
||||
|
||||
|
||||
def _load_prompts(
|
||||
*,
|
||||
prompts_root,
|
||||
skills_by_id: dict[str, SkillRecord],
|
||||
) -> tuple[dict[str, PromptRecord], list[PromptSummaryRecord], list[RegistryIssue]]:
|
||||
prompts_by_id: dict[str, PromptRecord] = {}
|
||||
prompt_summaries: list[PromptSummaryRecord] = []
|
||||
issues: list[RegistryIssue] = []
|
||||
discovered_prompt_docs: set[str] = set()
|
||||
if prompts_root.is_dir():
|
||||
for prompt_dir in sorted(prompts_root.iterdir(), key=lambda item: item.name):
|
||||
if not prompt_dir.is_dir():
|
||||
continue
|
||||
|
||||
prompt_dir_name = prompt_dir.name
|
||||
prompt_rel_root = PurePosixPath("prompts").joinpath(prompt_dir_name)
|
||||
prompt_doc_relpath = prompt_rel_root.joinpath("PROMPT.md").as_posix()
|
||||
prompt_doc_file = prompt_dir.joinpath("PROMPT.md")
|
||||
for prompt_dir in sorted(prompts_root.iterdir(), key=lambda item: item.name):
|
||||
if not prompt_dir.is_dir():
|
||||
continue
|
||||
|
||||
if not prompt_doc_file.is_file():
|
||||
continue
|
||||
prompt_dir_name = prompt_dir.name
|
||||
prompt_rel_root = PurePosixPath("prompts").joinpath(prompt_dir_name)
|
||||
prompt_doc_relpath = prompt_rel_root.joinpath("PROMPT.md").as_posix()
|
||||
prompt_doc_file = prompt_dir.joinpath("PROMPT.md")
|
||||
|
||||
discovered_prompt_docs.add(prompt_doc_relpath)
|
||||
if not prompt_doc_file.is_file():
|
||||
continue
|
||||
|
||||
prompt_markdown = prompt_doc_file.read_text(encoding="utf-8")
|
||||
try:
|
||||
raw_frontmatter, _ = _parse_frontmatter(
|
||||
prompt_markdown,
|
||||
discovered_prompt_docs.add(prompt_doc_relpath)
|
||||
|
||||
prompt_markdown = prompt_doc_file.read_text(encoding="utf-8")
|
||||
try:
|
||||
raw_frontmatter, _ = _parse_frontmatter(
|
||||
prompt_markdown,
|
||||
path=prompt_doc_relpath,
|
||||
)
|
||||
frontmatter = _validate_prompt_frontmatter(
|
||||
raw_frontmatter,
|
||||
prompt_dir_name=prompt_dir_name,
|
||||
)
|
||||
except (ValueError, ValidationError) as exc:
|
||||
issues.append(
|
||||
RegistryIssue(
|
||||
code="invalid_prompt_frontmatter",
|
||||
message=str(exc),
|
||||
skill_id=prompt_dir_name,
|
||||
path=prompt_doc_relpath,
|
||||
)
|
||||
frontmatter = _validate_prompt_frontmatter(
|
||||
raw_frontmatter,
|
||||
prompt_dir_name=prompt_dir_name,
|
||||
)
|
||||
except (ValueError, ValidationError) as exc:
|
||||
issues.append(
|
||||
RegistryIssue(
|
||||
code="invalid_prompt_frontmatter",
|
||||
message=str(exc),
|
||||
skill_id=prompt_dir_name,
|
||||
path=prompt_doc_relpath,
|
||||
hint="fix PROMPT.md YAML frontmatter to match the contract",
|
||||
)
|
||||
)
|
||||
continue
|
||||
|
||||
prompt_id = frontmatter.name
|
||||
if prompt_id in prompts_by_id:
|
||||
issues.append(
|
||||
RegistryIssue(
|
||||
code="duplicate_prompt_id",
|
||||
message="duplicate prompt id discovered",
|
||||
skill_id=prompt_id,
|
||||
path=prompt_doc_relpath,
|
||||
hint="ensure each prompt directory has a unique id",
|
||||
)
|
||||
)
|
||||
continue
|
||||
|
||||
if prompt_id in skills_by_id:
|
||||
issues.append(
|
||||
RegistryIssue(
|
||||
code="prompt_skill_id_collision",
|
||||
message="prompt id collides with an existing skill id",
|
||||
skill_id=prompt_id,
|
||||
path=prompt_doc_relpath,
|
||||
hint="use a unique prompt id that does not match a skill id",
|
||||
)
|
||||
)
|
||||
continue
|
||||
|
||||
prompt_record = PromptRecord(
|
||||
prompt_id=prompt_id,
|
||||
name=frontmatter.name,
|
||||
description=frontmatter.description,
|
||||
version=frontmatter.x_personal_mcp.version,
|
||||
tags=tuple(frontmatter.x_personal_mcp.tags),
|
||||
capabilities=tuple(frontmatter.x_personal_mcp.capabilities),
|
||||
arguments=frontmatter.x_personal_mcp.arguments,
|
||||
document_uri=f"resource://prompts/{prompt_id}/document",
|
||||
document_relpath=prompt_doc_relpath,
|
||||
document_content=prompt_markdown,
|
||||
)
|
||||
prompts_by_id[prompt_id] = prompt_record
|
||||
prompt_summaries.append(
|
||||
PromptSummaryRecord(
|
||||
prompt_id=prompt_record.prompt_id,
|
||||
name=prompt_record.name,
|
||||
description=prompt_record.description,
|
||||
tags=prompt_record.tags,
|
||||
capabilities=prompt_record.capabilities,
|
||||
document_uri=prompt_record.document_uri,
|
||||
version=prompt_record.version,
|
||||
hint="fix PROMPT.md YAML frontmatter to match the contract",
|
||||
)
|
||||
)
|
||||
continue
|
||||
|
||||
for relpath, prompt_file in _walk_markdown(
|
||||
prompts_root,
|
||||
prefix=PurePosixPath("prompts"),
|
||||
):
|
||||
if relpath in discovered_prompt_docs or relpath.endswith("/PROMPT.md"):
|
||||
continue
|
||||
|
||||
prompt_markdown = prompt_file.read_text(encoding="utf-8")
|
||||
prompt_id = _reference_id_from_filename(PurePosixPath(relpath).name)
|
||||
if prompt_id is None:
|
||||
continue
|
||||
if prompt_id in prompts_by_id:
|
||||
issues.append(
|
||||
RegistryIssue(
|
||||
code="duplicate_prompt_id",
|
||||
message="duplicate prompt id discovered",
|
||||
skill_id=prompt_id,
|
||||
path=relpath,
|
||||
hint="ensure each prompt id is unique",
|
||||
)
|
||||
)
|
||||
continue
|
||||
if prompt_id in skills_by_id:
|
||||
issues.append(
|
||||
RegistryIssue(
|
||||
code="prompt_skill_id_collision",
|
||||
message="prompt id collides with an existing skill id",
|
||||
skill_id=prompt_id,
|
||||
path=relpath,
|
||||
hint="use a unique prompt id that does not match a skill id",
|
||||
)
|
||||
)
|
||||
continue
|
||||
|
||||
prompt_record = PromptRecord(
|
||||
prompt_id=prompt_id,
|
||||
name=prompt_id,
|
||||
description=f"Legacy prompt loaded from docs/{relpath}",
|
||||
version="1.0.0",
|
||||
tags=("prompt", "legacy"),
|
||||
capabilities=(f"resource://prompts/{prompt_id}/document",),
|
||||
arguments={},
|
||||
document_uri=f"resource://prompts/{prompt_id}/document",
|
||||
document_relpath=relpath,
|
||||
document_content=prompt_markdown,
|
||||
)
|
||||
prompts_by_id[prompt_id] = prompt_record
|
||||
prompt_summaries.append(
|
||||
PromptSummaryRecord(
|
||||
prompt_id=prompt_record.prompt_id,
|
||||
name=prompt_record.name,
|
||||
description=prompt_record.description,
|
||||
tags=prompt_record.tags,
|
||||
capabilities=prompt_record.capabilities,
|
||||
document_uri=prompt_record.document_uri,
|
||||
version=prompt_record.version,
|
||||
prompt_id = frontmatter.name
|
||||
if prompt_id in prompts_by_id:
|
||||
issues.append(
|
||||
RegistryIssue(
|
||||
code="duplicate_prompt_id",
|
||||
message="duplicate prompt id discovered",
|
||||
skill_id=prompt_id,
|
||||
path=prompt_doc_relpath,
|
||||
hint="ensure each prompt directory has a unique id",
|
||||
)
|
||||
)
|
||||
continue
|
||||
|
||||
if prompt_id in skills_by_id:
|
||||
issues.append(
|
||||
RegistryIssue(
|
||||
code="prompt_skill_id_collision",
|
||||
message="prompt id collides with an existing skill id",
|
||||
skill_id=prompt_id,
|
||||
path=prompt_doc_relpath,
|
||||
hint="use a unique prompt id that does not match a skill id",
|
||||
)
|
||||
)
|
||||
continue
|
||||
|
||||
prompt_record = PromptRecord(
|
||||
prompt_id=prompt_id,
|
||||
name=frontmatter.name,
|
||||
description=frontmatter.description,
|
||||
version=frontmatter.x_personal_mcp.version,
|
||||
tags=tuple(frontmatter.x_personal_mcp.tags),
|
||||
capabilities=tuple(frontmatter.x_personal_mcp.capabilities),
|
||||
arguments=frontmatter.x_personal_mcp.arguments,
|
||||
document_uri=f"resource://prompts/{prompt_id}/document",
|
||||
document_relpath=prompt_doc_relpath,
|
||||
document_content=prompt_markdown,
|
||||
)
|
||||
prompts_by_id[prompt_id] = prompt_record
|
||||
prompt_summaries.append(
|
||||
PromptSummaryRecord(
|
||||
prompt_id=prompt_record.prompt_id,
|
||||
name=prompt_record.name,
|
||||
description=prompt_record.description,
|
||||
tags=prompt_record.tags,
|
||||
capabilities=prompt_record.capabilities,
|
||||
document_uri=prompt_record.document_uri,
|
||||
version=prompt_record.version,
|
||||
)
|
||||
)
|
||||
|
||||
for relpath, prompt_file in _walk_markdown(
|
||||
prompts_root,
|
||||
prefix=PurePosixPath("prompts"),
|
||||
):
|
||||
if relpath in discovered_prompt_docs or relpath.endswith("/PROMPT.md"):
|
||||
continue
|
||||
|
||||
prompt_markdown = prompt_file.read_text(encoding="utf-8")
|
||||
prompt_id = _reference_id_from_filename(PurePosixPath(relpath).name)
|
||||
if prompt_id is None:
|
||||
continue
|
||||
if prompt_id in prompts_by_id:
|
||||
issues.append(
|
||||
RegistryIssue(
|
||||
code="duplicate_prompt_id",
|
||||
message="duplicate prompt id discovered",
|
||||
skill_id=prompt_id,
|
||||
path=relpath,
|
||||
hint="ensure each prompt id is unique",
|
||||
)
|
||||
)
|
||||
continue
|
||||
if prompt_id in skills_by_id:
|
||||
issues.append(
|
||||
RegistryIssue(
|
||||
code="prompt_skill_id_collision",
|
||||
message="prompt id collides with an existing skill id",
|
||||
skill_id=prompt_id,
|
||||
path=relpath,
|
||||
hint="use a unique prompt id that does not match a skill id",
|
||||
)
|
||||
)
|
||||
continue
|
||||
|
||||
prompt_record = PromptRecord(
|
||||
prompt_id=prompt_id,
|
||||
name=prompt_id,
|
||||
description=f"Legacy prompt loaded from docs/{relpath}",
|
||||
version="1.0.0",
|
||||
tags=("prompt", "legacy"),
|
||||
capabilities=(f"resource://prompts/{prompt_id}/document",),
|
||||
arguments={},
|
||||
document_uri=f"resource://prompts/{prompt_id}/document",
|
||||
document_relpath=relpath,
|
||||
document_content=prompt_markdown,
|
||||
)
|
||||
prompts_by_id[prompt_id] = prompt_record
|
||||
prompt_summaries.append(
|
||||
PromptSummaryRecord(
|
||||
prompt_id=prompt_record.prompt_id,
|
||||
name=prompt_record.name,
|
||||
description=prompt_record.description,
|
||||
tags=prompt_record.tags,
|
||||
capabilities=prompt_record.capabilities,
|
||||
document_uri=prompt_record.document_uri,
|
||||
version=prompt_record.version,
|
||||
)
|
||||
)
|
||||
|
||||
return prompts_by_id, prompt_summaries, issues
|
||||
|
||||
|
||||
def _collect_relationship_and_uri_issues(
|
||||
*,
|
||||
skills_by_id: dict[str, SkillRecord],
|
||||
prompts_by_id: dict[str, PromptRecord],
|
||||
) -> list[RegistryIssue]:
|
||||
issues: list[RegistryIssue] = []
|
||||
|
||||
for skill_id, record in sorted(skills_by_id.items()):
|
||||
for dependency in record.depends_on:
|
||||
@@ -363,18 +389,84 @@ def load_docs_registry(
|
||||
seen_uris.add(uri)
|
||||
|
||||
for prompt_id, record in sorted(prompts_by_id.items()):
|
||||
for uri in [record.document_uri]:
|
||||
if uri in seen_uris:
|
||||
issues.append(
|
||||
RegistryIssue(
|
||||
code="duplicate_uri",
|
||||
message=f"duplicate resource URI generated: {uri}",
|
||||
skill_id=prompt_id,
|
||||
path=record.document_relpath,
|
||||
hint="ensure unique prompt ids",
|
||||
)
|
||||
uri = record.document_uri
|
||||
if uri in seen_uris:
|
||||
issues.append(
|
||||
RegistryIssue(
|
||||
code="duplicate_uri",
|
||||
message=f"duplicate resource URI generated: {uri}",
|
||||
skill_id=prompt_id,
|
||||
path=record.document_relpath,
|
||||
hint="ensure unique prompt ids",
|
||||
)
|
||||
seen_uris.add(uri)
|
||||
)
|
||||
seen_uris.add(uri)
|
||||
|
||||
return issues
|
||||
|
||||
|
||||
def load_docs_registry(
|
||||
*,
|
||||
package_anchor: str,
|
||||
docs_root: str = "docs",
|
||||
) -> DocsRegistry:
|
||||
docs_dir = files(package_anchor).joinpath(docs_root)
|
||||
|
||||
if not docs_dir.is_dir():
|
||||
raise DocsRegistryValidationError(
|
||||
[
|
||||
RegistryIssue(
|
||||
code="missing_docs_root",
|
||||
message="docs root directory does not exist",
|
||||
skill_id=None,
|
||||
path=docs_root,
|
||||
hint="configure docs_root to a valid packaged docs path",
|
||||
)
|
||||
]
|
||||
)
|
||||
|
||||
docs_markdown_by_path = {
|
||||
relpath: doc_file.read_text(encoding="utf-8") for relpath, doc_file in _walk_markdown(docs_dir)
|
||||
}
|
||||
|
||||
skills_root = docs_dir.joinpath("skills")
|
||||
if not skills_root.is_dir():
|
||||
raise DocsRegistryValidationError(
|
||||
[
|
||||
RegistryIssue(
|
||||
code="missing_skills_root",
|
||||
message="skills directory does not exist under docs root",
|
||||
skill_id=None,
|
||||
path=f"{docs_root}/skills",
|
||||
hint="ensure docs/skills is included in packaged docs",
|
||||
)
|
||||
]
|
||||
)
|
||||
|
||||
skills_by_id, summaries, skill_issues = _load_skills(
|
||||
docs_markdown_by_path=docs_markdown_by_path,
|
||||
skills_root=skills_root,
|
||||
)
|
||||
|
||||
prompts_by_id: dict[str, PromptRecord] = {}
|
||||
prompt_summaries: list[PromptSummaryRecord] = []
|
||||
prompt_issues: list[RegistryIssue] = []
|
||||
|
||||
prompts_root = docs_dir.joinpath("prompts")
|
||||
if prompts_root.is_dir():
|
||||
prompts_by_id, prompt_summaries, prompt_issues = _load_prompts(
|
||||
prompts_root=prompts_root,
|
||||
skills_by_id=skills_by_id,
|
||||
)
|
||||
|
||||
issues = [
|
||||
*skill_issues,
|
||||
*prompt_issues,
|
||||
*_collect_relationship_and_uri_issues(
|
||||
skills_by_id=skills_by_id,
|
||||
prompts_by_id=prompts_by_id,
|
||||
),
|
||||
]
|
||||
|
||||
if issues:
|
||||
raise DocsRegistryValidationError(issues)
|
||||
|
||||
@@ -1,3 +1,4 @@
|
||||
from ..skills.document_loader import _normalize_docs_path
|
||||
from .contracts import DocsRegistry
|
||||
|
||||
|
||||
|
||||
Reference in New Issue
Block a user