[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M699B8RZNMQJ42EYNCRYQ2FX":3,"skill-same:01M699B8RZNMQJ42EYNCRYQ2FX":73},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":38,"linkOnly":39,"markdownContent":40,"markdownBody":41,"frontMatter":42,"refName":43,"commitSha":44,"skillDirectory":45,"forksCount":46,"repositoryHtmlUrl":47,"repositoryHomepageUrl":48,"ownerHtmlUrl":49,"ownerBlogUrl":50,"ownerDescription":51,"updatedAt":52,"install":53,"resources":66},"01M699B8RZNMQJ42EYNCRYQ2FX","python-package-management",null,"Guide for managing packages in the Agent Framework Python monorepo, including creating new connector packages, versioning, and the lazy-loading pattern. Use this when adding, modifying, or releasing packages.","microsoft\u002Fagent-framework","microsoft","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F6154722?v=4","01M698VYDXZ6JW5WH3CRV650R8","Microsoft","01M6999R99RF2DY8AXYFS5HZRM",14676,"MIT","https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fagent-framework\u002Fblob\u002Fmain\u002Fpython\u002F.github\u002Fskills\u002Fpython-package-management\u002FSKILL.md",[],[19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37],"python","python package management","package","package management","management","when","lazy","lazy-loading","for","loading","pattern","Connector","use","guide","monorepo","versioning","framework","packages","agent framework","2026-11-10 10:12:30",false,"---\nname: python-package-management\ndescription: >\n  Guide for managing packages in the Agent Framework Python monorepo, including\n  creating new connector packages, versioning, and the lazy-loading pattern.\n  Use this when adding, modifying, or releasing packages.\n---\n\n# Python Package Management\n\n## Monorepo Structure\n\n```\npython\u002F\n├── pyproject.toml              # Root package (agent-framework)\n├── packages\u002F\n│   ├── core\u002F                   # agent-framework-core (main package)\n│   ├── foundry\u002F                # agent-framework-foundry\n│   ├── lab\u002F                    # Standalone uv project, excluded from the root workspace\n│   ├── anthropic\u002F              # agent-framework-anthropic\n│   └── ...                     # Other connector packages\n```\n\n- `agent-framework-core` contains core abstractions and OpenAI\u002FAzure OpenAI built-in\n- Provider packages extend core with specific integrations\n- Root `agent-framework` depends on `agent-framework-core[all]`. Despite the existing extra name, this is a\n  curated standard install set, not every package at a particular lifecycle stage.\n- Entries in `agent-framework-core[all]` carry lower and upper bounds. Workspace integrations use the version\n  released with the current Core cohort as their lower bound so pip does not backtrack across old cyclic\n  `core[all] -> integration -> core` combinations.\n- `packages\u002Flab` is intentionally excluded from the root uv workspace. It has its own\n  `pyproject.toml`, `uv.lock`, environment, CI matrix, and dependency updates so experimental\n  dependencies cannot constrain released provider packages. Lab resolves released Core and provider\n  distributions rather than local root-workspace packages.\n\n## Dependency Management\n\nUses [uv](https:\u002F\u002Fgithub.com\u002Fastral-sh\u002Fuv) for dependency management and\n[poethepoet](https:\u002F\u002Fgithub.com\u002Fnat-n\u002Fpoethepoet) for task automation.\n\n```bash\n# Full setup (venv + install + prek hooks)\nuv run poe setup\n\n# Install dependencies from lockfile (frozen resolution with prerelease policy)\nuv run poe install\n\n# Create venv with specific Python version\nuv run poe venv --python 3.12\n\n# Intentionally upgrade a specific dependency to reduce lockfile conflicts\nuv lock --upgrade-package \u003Cdependency-name> && uv run poe install\n\n# Refresh exact development and on-demand tool pins, lockfiles, and validation\nuv run poe upgrade-dev-dependencies\n\n# Release cuts: refresh uv.lock and probe changed packages at both bound extremes.\n# The release probe has a shared five-minute deadline.\nuv run poe validate-python-release --base-ref upstream\u002Fmain\n\n# Exhaustive test+typing matrix (slow; use for deliberate dependency-range work or CI)\nuv run poe validate-dependency-bounds-test\n# Defaults to --package \"*\"; scope locally whenever possible.\nuv run poe validate-dependency-bounds-test --package core\n\n# Then expand bounds for one dependency in the target package\nuv run poe validate-dependency-bounds-project --mode both --package core --dependency \"\u003Cdependency-name>\"\n\n# Repo-wide automation can reuse the same task\nuv run poe validate-dependency-bounds-project --mode upper --package \"*\"\n\n# Add a dependency to one project and run both validators for that project\u002Fdependency\nuv run poe add-dependency-and-validate-bounds --package core --dependency \"\u003Cdependency-spec>\"\n\n# Work on the standalone Lab project\ncd packages\u002Flab\nuv sync --all-extras --all-groups\nuv run poe test\n```\n\n### Dependency Bound Notes\n\n- Stable dependencies (`>=1.0`) should typically be bounded as `>=\u003Cknown-good>,\u003Cnext-major>`.\n- Prerelease (`dev`\u002F`a`\u002F`b`\u002F`rc`) and `\u003C1.0` dependencies should use hard bounds with an explicit upper cap (avoid open-ended ranges).\n- For `\u003C1.0` dependencies, prefer the broadest validated range the package can really support. That may be a patch line, a minor line, or multiple minor lines when checks\u002Ftests show the broader lane is compatible.\n- Prefer supporting multiple majors when practical; if APIs diverge across supported majors, use version-conditional imports\u002Fpaths.\n- For release-only version, lifecycle, pin, and internal-floor edits, use `validate-python-release`. It refreshes\n  `uv.lock`, finds changed package metadata relative to the selected main ref, and runs the changed packages'\n  published runtime dependencies and non-development extras through lock-independent `lowest-direct` and `highest`\n  import probes on the minimum Python minor supported by each package's internal editable closure. The probes run\n  concurrently under one 300-second deadline; pass `--python` only when an explicit interpreter override is needed.\n- For deliberate external dependency-range changes, use\n  `validate-dependency-bounds-project --mode both` for the target package\u002Fdependency to find and validate the actual\n  minimum and maximum constraints. Scope the exhaustive `validate-dependency-bounds-test` matrix to affected\n  packages during local iteration; reserve the workspace-wide form for CI or an intentional full audit. The same\n  project task can drive repo-wide upper-bound automation by using `--package \"*\"` and omitting `--dependency`.\n- Prefer targeted lock updates with `uv lock --upgrade-package \u003Cdependency-name>` to reduce `uv.lock` merge conflicts.\n- Use `add-dependency-and-validate-bounds` for package-scoped dependency additions plus bound validation in one command.\n- Keep the test framework and Poe in the root `dev` group across Python\n  versions. Put script-only libraries in `tool-runtime`. Put tools used by\n  only one Python version in a named `tool-*` group or pinned\n  `tooling\u002Frequirements-*.txt` file, and add the smallest relevant group or\n  bundle to each command. Keep source\u002Ftype-check support that is part of a\n  package's import surface in a package dependency group. Put package-specific\n  test fixtures in a `test` group.\n- Use `upgrade-dev-dependencies` for repo-wide development and tool dependency\n  refreshes; it repins exact dependencies across development groups and\n  `tooling\u002Frequirements-*.txt`, refreshes `uv.lock`, and reruns `check` and\n  `typing`.\n- Upgrade the root uv version manually as one coordinated change across the\n  `dev` requirement, `[tool.uv].required-version`, setup workflows, and the\n  uv pre-commit revision. Automated dependency updates intentionally ignore\n  the root uv pin.\n- Run Lab dependency and validation commands from `packages\u002Flab`; root workspace tasks deliberately do not\n  discover or update the Lab project.\n\n## Meta-package Membership\n\nPackage lifecycle and membership in the root `agent-framework` meta-package are independent decisions. A package\nstarts outside `agent-framework-core[all]`, and promotion to `beta`, `rc`, or `released` does not add it\nautomatically.\n\nAdding a package to or removing a package from the standard install set requires an agreed GitHub issue and an\nexplicit core-team decision. For additions, consider:\n\n- whether the package is useful across common Agent Framework workloads rather than for one specialized storage\n  backend, hosting adapter, channel, or deployment target;\n- whether its dependency footprint is proportionate and compatible with the meta-package's supported Python and\n  platform matrix, especially for native, heavyweight, or conflict-prone dependencies;\n- whether it belongs in the documented standard getting-started experience; and\n- whether including one package would create an expectation that an open-ended family of alternative backends or\n  adapters must also be included.\n\nDefault to a separate install when the case is unclear. Existing membership is not precedent for every package in\nthe same lifecycle stage or category.\n\nFor removals, the issue must also assess compatibility for existing `pip install agent-framework` and\n`agent-framework-core[all]` users, document any migration or replacement path, determine whether the change is\nbreaking, and identify the required Core and root meta-package version changes. Do not treat removal as ordinary\ndependency cleanup.\n\nTrack the decision in the `Meta package` column of `python\u002FPACKAGE_STATUS.md`. When membership changes, update\n`packages\u002Fcore\u002Fpyproject.toml` `[project.optional-dependencies].all`, the Core and root package versions as required,\nand the status table in the same change.\n\n## Lazy Loading Pattern\n\n### Root core API\n\nThe root `agent_framework` package is a lazy public API surface:\n\n- Runtime exports live in `packages\u002Fcore\u002Fagent_framework\u002F__init__.py`.\n- Typing\u002Feditor exports live in `packages\u002Fcore\u002Fagent_framework\u002F__init__.pyi`.\n- Add or move root exports in `_LAZY_MODULE_EXPORTS`, keep the explicit runtime `__all__` in sync, and add the same\n  symbol to the `.pyi` file.\n- Keep deprecation behavior in the owning module (for example, a module-level `__getattr__` that warns and returns\n  the deprecated alias). Do not add one-off deprecated-symbol branches to root `__getattr__`.\n- Validate root API changes with `uv run poe syntax -P core`, `uv run poe pyright -P core`, and import smoke tests\n  for both `from agent_framework import \u003Csymbol>` and `from agent_framework import *`.\n\n### Provider namespaces\n\nProvider folders in core use `__getattr__` to lazy load from connector packages:\n\n```python\n# In agent_framework\u002Ffoundry\u002F__init__.py\n_IMPORTS: dict[str, tuple[str, str]] = {\n    \"FoundryChatClient\": (\"agent_framework_foundry\", \"agent-framework-foundry\"),\n}\n\ndef __getattr__(name: str) -> Any:\n    if name in _IMPORTS:\n        import_path, package_name = _IMPORTS[name]\n        try:\n            return getattr(importlib.import_module(import_path), name)\n        except ModuleNotFoundError as exc:\n            raise ModuleNotFoundError(\n                f\"The package {package_name} is required to use `{name}`. \"\n                f\"Install it with: pip install {package_name}\"\n            ) from exc\n```\n\n### Namespace compatibility\n\nThe core namespace module and the objects it re-exports have different compatibility rules:\n\n- While an owning package is `alpha`, `beta`, or `rc`, imports for objects from that package may be added, removed,\n  or moved between approved lazy-loading namespaces by explicit core-team decision. Those object paths follow the\n  owning package's prerelease lifecycle.\n- Do not remove a core namespace module unless it is empty after the change and is used exclusively by packages\n  that have not reached `released`.\n- If a namespace exposes objects from any released package, keep the namespace. Changing import paths for released\n  objects requires a deprecated forwarding path where feasible; otherwise the agreed issue must include migration\n  guidance, explicit breaking-change acknowledgement, and the required Core and root meta-package version changes.\n\n## Adding a New Connector Package\n\n**Important:** Do not create a new package unless approved by the core team.\n\nEvery new package starts as `alpha`.\n\n### Alpha package checklist\n\n1. Create directory under `packages\u002F` (e.g., `packages\u002Fmy-connector\u002F`)\n2. Add the package to `tool.uv.sources` in root `pyproject.toml`\n3. Set the package version to the alpha pattern: `1.0.0a\u003Cdate>`\n4. Set the package classifier to `Development Status :: 3 - Alpha`\n5. Include samples inside the package (e.g., `packages\u002Fmy-connector\u002Fsamples\u002F`)\n6. Do **NOT** add to `[all]` extra in `packages\u002Fcore\u002Fpyproject.toml`\n7. Do **NOT** create lazy loading in core yet\n8. Add the package to `python\u002FPACKAGE_STATUS.md` with lifecycle state `alpha` and meta-package membership `no`,\n   and keep that file updated when packages are added, removed, renamed, promoted, or have their membership\n   changed. If the package exposes individually staged APIs, keep the feature list there current too.\n9. Keep the `Typing :: Typed` classifier and the `py.typed` marker in step. The classifier alone is a\n   promise PEP 561 keeps through the `py.typed` file inside the package directory, and that file is\n   the signal checkers act on: a package that declares `Typing :: Typed` without shipping it is not\n   treated as typed, so its API falls back to whatever a checker does with an untyped import\n   (`Any`, `Unknown`, or a missing-stub diagnostic) instead of the inline annotations. Check both,\n   not one:\n\n   ```bash\n   # every package that claims to be typed must ship the marker\n   cd python\u002Fpackages\n   for d in *\u002F; do\n     d=${d%\u002F}\n     [ -f \"$d\u002Fpyproject.toml\" ] || continue\n     if grep -q \"Typing :: Typed\" \"$d\u002Fpyproject.toml\" && ! ls \"$d\"\u002Fagent_framework*\u002Fpy.typed >\u002Fdev\u002Fnull 2>&1; then\n       echo \"missing py.typed: $d\"\n     fi\n   done\n   ```\n\n   Note that `py.typed` only makes annotations that are already there visible; it adds none, so a\n   package that has not been type-checked may surface further errors once the marker lands.\n\nRecommended dependency workflow during connector implementation:\n\n1. Add the dependency to the target package:\n   `uv run poe add-dependency-to-project --package core --dependency \"\u003Cdependency-spec>\"`\n2. Implement connector code and tests.\n3. Validate dependency bounds for that package\u002Fdependency:\n   `uv run poe validate-dependency-bounds-project --mode both --package core --dependency \"\u003Cdependency-name>\"`\n4. If the package has meaningful tests\u002Fchecks that validate dependency compatibility, you can use the add + validation flow in one command:\n   `uv run poe add-dependency-and-validate-bounds --package core --dependency \"\u003Cdependency-spec>\"`\n   If compatibility checks are not in place yet, add the dependency first, then implement tests before running bound validation.\n\n### Promotion path\n\nCreate a GitHub issue for every package promotion and get core-team agreement before implementing the promotion.\nThe issue must record:\n\n- the current and target lifecycle stages;\n- whether the package should be included in the root meta-package, with rationale based on the criteria above; and\n- the lazy-loading namespace decision for the transition: `alpha` -> `beta` selects whether and where to expose the\n  package, while `beta` -> `rc` and `rc` -> `released` preserve the existing namespace by default and may change\n  the package's prerelease object imports only by explicit core-team decision and under the namespace compatibility\n  rules above.\n\nRecord an explicit `no` or `none` when the package should remain outside the meta-package or should not have a\nlazy-loading namespace. Do not infer either decision from the target lifecycle stage.\n\nPromotion work is not isolated to the package being promoted. If a promotion changes dependency\nmetadata for downstream packages, also update the dependent packages' own versions so they publish\nnew metadata alongside the promoted dependency bounds.\nApply the internal package dependency update rules from the versioning section below during\npromotions as well as standalone version update work.\n\n#### Alpha -> Beta\n\nMove a package to `beta` when it is stable enough for broader prerelease use.\n\n1. Update the package version to the beta pattern: `1.0.0b\u003Cdate>`\n2. Update the classifier to `Development Status :: 4 - Beta`\n3. Move samples to the root `samples\u002F` tree and remove package-local samples\n4. Apply the lazy-loading decision recorded in the promotion issue, creating or updating the approved namespace\n   when applicable\n5. Apply the meta-package decision recorded in the promotion issue. The default is to remain excluded; add the\n   package to `[all]` only when inclusion is explicitly approved.\n6. Update the lifecycle state and meta-package membership in `python\u002FPACKAGE_STATUS.md`\n\nAfter `alpha`, there should be no samples left inside a package folder.\n\n#### Beta -> RC\n\nMove a package to `rc` when its API is close to the final released shape.\n\n1. Update the package version to the release-candidate pattern: `1.0.0rc\u003Cnumber>`\n2. Keep the classifier at `Development Status :: 4 - Beta` because PyPI does not have a separate\n   release-candidate classifier\n3. Apply the meta-package decision recorded in the promotion issue, preserving current membership when unchanged\n   or using the membership-change process above when an addition or removal is explicitly approved; the stage\n   transition alone does not add it to `core[all]`\n4. Preserve the existing lazy-loading namespace by default. Changes to the package's prerelease object imports are\n   discouraged and require explicit core-team approval recorded in the promotion issue; apply the namespace\n   compatibility rules above\n5. Keep samples only in the root `samples\u002F` tree\n6. Update `python\u002FPACKAGE_STATUS.md` to show the package as `rc` and keep its membership value current\n\n#### RC -> Released\n\nMove a package to `released` when it no longer carries a prerelease qualifier.\n\n1. Update the package version to the stable pattern: `1.0.0`\n2. Update the classifier to `Development Status :: 5 - Production\u002FStable`\n3. Apply the meta-package decision recorded in the promotion issue, preserving current membership when unchanged\n   or using the membership-change process above when an addition or removal is explicitly approved; release does\n   not imply inclusion in `core[all]`\n4. Preserve the lazy-loading namespace used at RC by default. Any final changes to the package's prerelease object\n   imports require explicit core-team approval recorded in the promotion issue; apply the namespace compatibility\n   rules above\n5. Keep samples only in the root `samples\u002F` tree\n6. Update `python\u002FPACKAGE_STATUS.md` to show the package as `released` and keep its membership value current\n7. Update all `README.md` files that install that package with\n   `pip install agent-framework-... --pre` so they use `pip install agent-framework-...` without\n   the `--pre` suffix\n\n## Versioning\n\n### Internal package dependency updates\n\n- If package A depends on package B within this repository, only update package A's dependency\n  declaration when the work on package B actually affects package A.\n- If package A does not need anything from the package B change, leave package A's dependency\n  declaration unchanged.\n- If package A does need something from the package B change, update package A's dependency\n  declaration to the version or versioning scheme that matches what package A now requires.\n- If package B is promoted to a different lifecycle stage, update package A's dependency\n  declaration to the new versioning scheme for package B even when the only change is the stage\n  transition itself.\n- When Core releases, advance each workspace package floor in `agent-framework-core[all]` to that package's\n  co-released version, including unchanged packages at their existing version. Keep an upper bound appropriate\n  for the package's current major version. This release-cohort metadata is independent of whether the package\n  itself needed a version bump.\n- Use this guidance both for ordinary version updates and for package promotion work.\n\n- All non-core packages declare a lower bound on `agent-framework-core`\n- When core version bumps with breaking changes, update the lower bound in all packages\n- Non-core packages version independently; only raise core bound when using new core APIs\n- If promoting a package changes a dependent package's published dependency metadata, bump the\n  dependent package's own version in the correct lifecycle pattern for its current stage\n- Lifecycle version patterns:\n  - `alpha`: `1.0.0a\u003Cdate>` where `\u003Cdate>` is the current Pacific (US west coast) `YYMMDD`\n  - `beta`: `1.0.0b\u003Cdate>` where `\u003Cdate>` is the current Pacific (US west coast) `YYMMDD`\n  - `rc`: `1.0.0rc\u003Cnumber>` where `\u003Cnumber>` increments only when the package has changes\n  - `released`: `X.Y.Z` using semver per package\n- For alpha\u002Fbeta date stamps, use the current Pacific date as the cutoff, not UTC and not the user's local\n  timezone. Same-Pacific-day re-cuts use a `.postN` suffix. Honor an explicit user-provided date over this\n  default.\n- Keep the `Development Status` classifier in `pyproject.toml` aligned with the lifecycle stage:\n  - `alpha` -> `Development Status :: 3 - Alpha`\n  - `beta` -> `Development Status :: 4 - Beta`\n  - `rc` -> `Development Status :: 4 - Beta`\n  - `released` -> `Development Status :: 5 - Production\u002FStable`\n- See the PyPI classifier list for the available classifier values:\n  `https:\u002F\u002Fpypi.org\u002Fclassifiers\u002F`\n\n## Installation Options\n\n```bash\npip install agent-framework-core          # Core only\npip install agent-framework-core[all]     # Core + curated standard integrations\npip install agent-framework               # Same standard install set as core[all]\npip install agent-framework-foundry       # Specific connector (pulls in core)\n```\n\n## Maintaining Documentation\n\nWhen changing a package, check if its `AGENTS.md` needs updates:\n- Adding\u002Fremoving\u002Frenaming public classes or functions\n- Changing the package's purpose or architecture\n- Modifying import paths or usage patterns\n\nKeep `python\u002FPACKAGE_STATUS.md` updated when:\n- A package is added, removed, renamed, or promoted between lifecycle stages\n- A package is added to or removed from the root meta-package\n- A package starts or stops exposing individually staged experimental or release-candidate APIs\n\nWhen a package adds, removes, or renames environment variables, update the related documentation in the same\nchange:\n- The package's `README.md` for package-level configuration\u002Fenv var guidance\n- `samples\u002FREADME.md` if the package is included in `packages\u002Fcore\u002Fpyproject.toml` `[all]` and the env var is\n  part of the consolidated package env-var inventory\n- Any affected sample\u002Fpackage-local `.env.example`, `.env.template`, or sample README files when sample setup\n  changes alongside the package\n","# Python Package Management\n\n## Monorepo Structure\n\n```\npython\u002F\n├── pyproject.toml              # Root package (agent-framework)\n├── packages\u002F\n│   ├── core\u002F                   # agent-framework-core (main package)\n│   ├── foundry\u002F                # agent-framework-foundry\n│   ├── lab\u002F                    # Standalone uv project, excluded from the root workspace\n│   ├── anthropic\u002F              # agent-framework-anthropic\n│   └── ...                     # Other connector packages\n```\n\n- `agent-framework-core` contains core abstractions and OpenAI\u002FAzure OpenAI built-in\n- Provider packages extend core with specific integrations\n- Root `agent-framework` depends on `agent-framework-core[all]`. Despite the existing extra name, this is a\n  curated standard install set, not every package at a particular lifecycle stage.\n- Entries in `agent-framework-core[all]` carry lower and upper bounds. Workspace integrations use the version\n  released with the current Core cohort as their lower bound so pip does not backtrack across old cyclic\n  `core[all] -> integration -> core` combinations.\n- `packages\u002Flab` is intentionally excluded from the root uv workspace. It has its own\n  `pyproject.toml`, `uv.lock`, environment, CI matrix, and dependency updates so experimental\n  dependencies cannot constrain released provider packages. Lab resolves released Core and provider\n  distributions rather than local root-workspace packages.\n\n## Dependency Management\n\nUses [uv](https:\u002F\u002Fgithub.com\u002Fastral-sh\u002Fuv) for dependency management and\n[poethepoet](https:\u002F\u002Fgithub.com\u002Fnat-n\u002Fpoethepoet) for task automation.\n\n```bash\n# Full setup (venv + install + prek hooks)\nuv run poe setup\n\n# Install dependencies from lockfile (frozen resolution with prerelease policy)\nuv run poe install\n\n# Create venv with specific Python version\nuv run poe venv --python 3.12\n\n# Intentionally upgrade a specific dependency to reduce lockfile conflicts\nuv lock --upgrade-package \u003Cdependency-name> && uv run poe install\n\n# Refresh exact development and on-demand tool pins, lockfiles, and validation\nuv run poe upgrade-dev-dependencies\n\n# Release cuts: refresh uv.lock and probe changed packages at both bound extremes.\n# The release probe has a shared five-minute deadline.\nuv run poe validate-python-release --base-ref upstream\u002Fmain\n\n# Exhaustive test+typing matrix (slow; use for deliberate dependency-range work or CI)\nuv run poe validate-dependency-bounds-test\n# Defaults to --package \"*\"; scope locally whenever possible.\nuv run poe validate-dependency-bounds-test --package core\n\n# Then expand bounds for one dependency in the target package\nuv run poe validate-dependency-bounds-project --mode both --package core --dependency \"\u003Cdependency-name>\"\n\n# Repo-wide automation can reuse the same task\nuv run poe validate-dependency-bounds-project --mode upper --package \"*\"\n\n# Add a dependency to one project and run both validators for that project\u002Fdependency\nuv run poe add-dependency-and-validate-bounds --package core --dependency \"\u003Cdependency-spec>\"\n\n# Work on the standalone Lab project\ncd packages\u002Flab\nuv sync --all-extras --all-groups\nuv run poe test\n```\n\n### Dependency Bound Notes\n\n- Stable dependencies (`>=1.0`) should typically be bounded as `>=\u003Cknown-good>,\u003Cnext-major>`.\n- Prerelease (`dev`\u002F`a`\u002F`b`\u002F`rc`) and `\u003C1.0` dependencies should use hard bounds with an explicit upper cap (avoid open-ended ranges).\n- For `\u003C1.0` dependencies, prefer the broadest validated range the package can really support. That may be a patch line, a minor line, or multiple minor lines when checks\u002Ftests show the broader lane is compatible.\n- Prefer supporting multiple majors when practical; if APIs diverge across supported majors, use version-conditional imports\u002Fpaths.\n- For release-only version, lifecycle, pin, and internal-floor edits, use `validate-python-release`. It refreshes\n  `uv.lock`, finds changed package metadata relative to the selected main ref, and runs the changed packages'\n  published runtime dependencies and non-development extras through lock-independent `lowest-direct` and `highest`\n  import probes on the minimum Python minor supported by each package's internal editable closure. The probes run\n  concurrently under one 300-second deadline; pass `--python` only when an explicit interpreter override is needed.\n- For deliberate external dependency-range changes, use\n  `validate-dependency-bounds-project --mode both` for the target package\u002Fdependency to find and validate the actual\n  minimum and maximum constraints. Scope the exhaustive `validate-dependency-bounds-test` matrix to affected\n  packages during local iteration; reserve the workspace-wide form for CI or an intentional full audit. The same\n  project task can drive repo-wide upper-bound automation by using `--package \"*\"` and omitting `--dependency`.\n- Prefer targeted lock updates with `uv lock --upgrade-package \u003Cdependency-name>` to reduce `uv.lock` merge conflicts.\n- Use `add-dependency-and-validate-bounds` for package-scoped dependency additions plus bound validation in one command.\n- Keep the test framework and Poe in the root `dev` group across Python\n  versions. Put script-only libraries in `tool-runtime`. Put tools used by\n  only one Python version in a named `tool-*` group or pinned\n  `tooling\u002Frequirements-*.txt` file, and add the smallest relevant group or\n  bundle to each command. Keep source\u002Ftype-check support that is part of a\n  package's import surface in a package dependency group. Put package-specific\n  test fixtures in a `test` group.\n- Use `upgrade-dev-dependencies` for repo-wide development and tool dependency\n  refreshes; it repins exact dependencies across development groups and\n  `tooling\u002Frequirements-*.txt`, refreshes `uv.lock`, and reruns `check` and\n  `typing`.\n- Upgrade the root uv version manually as one coordinated change across the\n  `dev` requirement, `[tool.uv].required-version`, setup workflows, and the\n  uv pre-commit revision. Automated dependency updates intentionally ignore\n  the root uv pin.\n- Run Lab dependency and validation commands from `packages\u002Flab`; root workspace tasks deliberately do not\n  discover or update the Lab project.\n\n## Meta-package Membership\n\nPackage lifecycle and membership in the root `agent-framework` meta-package are independent decisions. A package\nstarts outside `agent-framework-core[all]`, and promotion to `beta`, `rc`, or `released` does not add it\nautomatically.\n\nAdding a package to or removing a package from the standard install set requires an agreed GitHub issue and an\nexplicit core-team decision. For additions, consider:\n\n- whether the package is useful across common Agent Framework workloads rather than for one specialized storage\n  backend, hosting adapter, channel, or deployment target;\n- whether its dependency footprint is proportionate and compatible with the meta-package's supported Python and\n  platform matrix, especially for native, heavyweight, or conflict-prone dependencies;\n- whether it belongs in the documented standard getting-started experience; and\n- whether including one package would create an expectation that an open-ended family of alternative backends or\n  adapters must also be included.\n\nDefault to a separate install when the case is unclear. Existing membership is not precedent for every package in\nthe same lifecycle stage or category.\n\nFor removals, the issue must also assess compatibility for existing `pip install agent-framework` and\n`agent-framework-core[all]` users, document any migration or replacement path, determine whether the change is\nbreaking, and identify the required Core and root meta-package version changes. Do not treat removal as ordinary\ndependency cleanup.\n\nTrack the decision in the `Meta package` column of `python\u002FPACKAGE_STATUS.md`. When membership changes, update\n`packages\u002Fcore\u002Fpyproject.toml` `[project.optional-dependencies].all`, the Core and root package versions as required,\nand the status table in the same change.\n\n## Lazy Loading Pattern\n\n### Root core API\n\nThe root `agent_framework` package is a lazy public API surface:\n\n- Runtime exports live in `packages\u002Fcore\u002Fagent_framework\u002F__init__.py`.\n- Typing\u002Feditor exports live in `packages\u002Fcore\u002Fagent_framework\u002F__init__.pyi`.\n- Add or move root exports in `_LAZY_MODULE_EXPORTS`, keep the explicit runtime `__all__` in sync, and add the same\n  symbol to the `.pyi` file.\n- Keep deprecation behavior in the owning module (for example, a module-level `__getattr__` that warns and returns\n  the deprecated alias). Do not add one-off deprecated-symbol branches to root `__getattr__`.\n- Validate root API changes with `uv run poe syntax -P core`, `uv run poe pyright -P core`, and import smoke tests\n  for both `from agent_framework import \u003Csymbol>` and `from agent_framework import *`.\n\n### Provider namespaces\n\nProvider folders in core use `__getattr__` to lazy load from connector packages:\n\n```python\n# In agent_framework\u002Ffoundry\u002F__init__.py\n_IMPORTS: dict[str, tuple[str, str]] = {\n    \"FoundryChatClient\": (\"agent_framework_foundry\", \"agent-framework-foundry\"),\n}\n\ndef __getattr__(name: str) -> Any:\n    if name in _IMPORTS:\n        import_path, package_name = _IMPORTS[name]\n        try:\n            return getattr(importlib.import_module(import_path), name)\n        except ModuleNotFoundError as exc:\n            raise ModuleNotFoundError(\n                f\"The package {package_name} is required to use `{name}`. \"\n                f\"Install it with: pip install {package_name}\"\n            ) from exc\n```\n\n### Namespace compatibility\n\nThe core namespace module and the objects it re-exports have different compatibility rules:\n\n- While an owning package is `alpha`, `beta`, or `rc`, imports for objects from that package may be added, removed,\n  or moved between approved lazy-loading namespaces by explicit core-team decision. Those object paths follow the\n  owning package's prerelease lifecycle.\n- Do not remove a core namespace module unless it is empty after the change and is used exclusively by packages\n  that have not reached `released`.\n- If a namespace exposes objects from any released package, keep the namespace. Changing import paths for released\n  objects requires a deprecated forwarding path where feasible; otherwise the agreed issue must include migration\n  guidance, explicit breaking-change acknowledgement, and the required Core and root meta-package version changes.\n\n## Adding a New Connector Package\n\n**Important:** Do not create a new package unless approved by the core team.\n\nEvery new package starts as `alpha`.\n\n### Alpha package checklist\n\n1. Create directory under `packages\u002F` (e.g., `packages\u002Fmy-connector\u002F`)\n2. Add the package to `tool.uv.sources` in root `pyproject.toml`\n3. Set the package version to the alpha pattern: `1.0.0a\u003Cdate>`\n4. Set the package classifier to `Development Status :: 3 - Alpha`\n5. Include samples inside the package (e.g., `packages\u002Fmy-connector\u002Fsamples\u002F`)\n6. Do **NOT** add to `[all]` extra in `packages\u002Fcore\u002Fpyproject.toml`\n7. Do **NOT** create lazy loading in core yet\n8. Add the package to `python\u002FPACKAGE_STATUS.md` with lifecycle state `alpha` and meta-package membership `no`,\n   and keep that file updated when packages are added, removed, renamed, promoted, or have their membership\n   changed. If the package exposes individually staged APIs, keep the feature list there current too.\n9. Keep the `Typing :: Typed` classifier and the `py.typed` marker in step. The classifier alone is a\n   promise PEP 561 keeps through the `py.typed` file inside the package directory, and that file is\n   the signal checkers act on: a package that declares `Typing :: Typed` without shipping it is not\n   treated as typed, so its API falls back to whatever a checker does with an untyped import\n   (`Any`, `Unknown`, or a missing-stub diagnostic) instead of the inline annotations. Check both,\n   not one:\n\n   ```bash\n   # every package that claims to be typed must ship the marker\n   cd python\u002Fpackages\n   for d in *\u002F; do\n     d=${d%\u002F}\n     [ -f \"$d\u002Fpyproject.toml\" ] || continue\n     if grep -q \"Typing :: Typed\" \"$d\u002Fpyproject.toml\" && ! ls \"$d\"\u002Fagent_framework*\u002Fpy.typed >\u002Fdev\u002Fnull 2>&1; then\n       echo \"missing py.typed: $d\"\n     fi\n   done\n   ```\n\n   Note that `py.typed` only makes annotations that are already there visible; it adds none, so a\n   package that has not been type-checked may surface further errors once the marker lands.\n\nRecommended dependency workflow during connector implementation:\n\n1. Add the dependency to the target package:\n   `uv run poe add-dependency-to-project --package core --dependency \"\u003Cdependency-spec>\"`\n2. Implement connector code and tests.\n3. Validate dependency bounds for that package\u002Fdependency:\n   `uv run poe validate-dependency-bounds-project --mode both --package core --dependency \"\u003Cdependency-name>\"`\n4. If the package has meaningful tests\u002Fchecks that validate dependency compatibility, you can use the add + validation flow in one command:\n   `uv run poe add-dependency-and-validate-bounds --package core --dependency \"\u003Cdependency-spec>\"`\n   If compatibility checks are not in place yet, add the dependency first, then implement tests before running bound validation.\n\n### Promotion path\n\nCreate a GitHub issue for every package promotion and get core-team agreement before implementing the promotion.\nThe issue must record:\n\n- the current and target lifecycle stages;\n- whether the package should be included in the root meta-package, with rationale based on the criteria above; and\n- the lazy-loading namespace decision for the transition: `alpha` -> `beta` selects whether and where to expose the\n  package, while `beta` -> `rc` and `rc` -> `released` preserve the existing namespace by default and may change\n  the package's prerelease object imports only by explicit core-team decision and under the namespace compatibility\n  rules above.\n\nRecord an explicit `no` or `none` when the package should remain outside the meta-package or should not have a\nlazy-loading namespace. Do not infer either decision from the target lifecycle stage.\n\nPromotion work is not isolated to the package being promoted. If a promotion changes dependency\nmetadata for downstream packages, also update the dependent packages' own versions so they publish\nnew metadata alongside the promoted dependency bounds.\nApply the internal package dependency update rules from the versioning section below during\npromotions as well as standalone version update work.\n\n#### Alpha -> Beta\n\nMove a package to `beta` when it is stable enough for broader prerelease use.\n\n1. Update the package version to the beta pattern: `1.0.0b\u003Cdate>`\n2. Update the classifier to `Development Status :: 4 - Beta`\n3. Move samples to the root `samples\u002F` tree and remove package-local samples\n4. Apply the lazy-loading decision recorded in the promotion issue, creating or updating the approved namespace\n   when applicable\n5. Apply the meta-package decision recorded in the promotion issue. The default is to remain excluded; add the\n   package to `[all]` only when inclusion is explicitly approved.\n6. Update the lifecycle state and meta-package membership in `python\u002FPACKAGE_STATUS.md`\n\nAfter `alpha`, there should be no samples left inside a package folder.\n\n#### Beta -> RC\n\nMove a package to `rc` when its API is close to the final released shape.\n\n1. Update the package version to the release-candidate pattern: `1.0.0rc\u003Cnumber>`\n2. Keep the classifier at `Development Status :: 4 - Beta` because PyPI does not have a separate\n   release-candidate classifier\n3. Apply the meta-package decision recorded in the promotion issue, preserving current membership when unchanged\n   or using the membership-change process above when an addition or removal is explicitly approved; the stage\n   transition alone does not add it to `core[all]`\n4. Preserve the existing lazy-loading namespace by default. Changes to the package's prerelease object imports are\n   discouraged and require explicit core-team approval recorded in the promotion issue; apply the namespace\n   compatibility rules above\n5. Keep samples only in the root `samples\u002F` tree\n6. Update `python\u002FPACKAGE_STATUS.md` to show the package as `rc` and keep its membership value current\n\n#### RC -> Released\n\nMove a package to `released` when it no longer carries a prerelease qualifier.\n\n1. Update the package version to the stable pattern: `1.0.0`\n2. Update the classifier to `Development Status :: 5 - Production\u002FStable`\n3. Apply the meta-package decision recorded in the promotion issue, preserving current membership when unchanged\n   or using the membership-change process above when an addition or removal is explicitly approved; release does\n   not imply inclusion in `core[all]`\n4. Preserve the lazy-loading namespace used at RC by default. Any final changes to the package's prerelease object\n   imports require explicit core-team approval recorded in the promotion issue; apply the namespace compatibility\n   rules above\n5. Keep samples only in the root `samples\u002F` tree\n6. Update `python\u002FPACKAGE_STATUS.md` to show the package as `released` and keep its membership value current\n7. Update all `README.md` files that install that package with\n   `pip install agent-framework-... --pre` so they use `pip install agent-framework-...` without\n   the `--pre` suffix\n\n## Versioning\n\n### Internal package dependency updates\n\n- If package A depends on package B within this repository, only update package A's dependency\n  declaration when the work on package B actually affects package A.\n- If package A does not need anything from the package B change, leave package A's dependency\n  declaration unchanged.\n- If package A does need something from the package B change, update package A's dependency\n  declaration to the version or versioning scheme that matches what package A now requires.\n- If package B is promoted to a different lifecycle stage, update package A's dependency\n  declaration to the new versioning scheme for package B even when the only change is the stage\n  transition itself.\n- When Core releases, advance each workspace package floor in `agent-framework-core[all]` to that package's\n  co-released version, including unchanged packages at their existing version. Keep an upper bound appropriate\n  for the package's current major version. This release-cohort metadata is independent of whether the package\n  itself needed a version bump.\n- Use this guidance both for ordinary version updates and for package promotion work.\n\n- All non-core packages declare a lower bound on `agent-framework-core`\n- When core version bumps with breaking changes, update the lower bound in all packages\n- Non-core packages version independently; only raise core bound when using new core APIs\n- If promoting a package changes a dependent package's published dependency metadata, bump the\n  dependent package's own version in the correct lifecycle pattern for its current stage\n- Lifecycle version patterns:\n  - `alpha`: `1.0.0a\u003Cdate>` where `\u003Cdate>` is the current Pacific (US west coast) `YYMMDD`\n  - `beta`: `1.0.0b\u003Cdate>` where `\u003Cdate>` is the current Pacific (US west coast) `YYMMDD`\n  - `rc`: `1.0.0rc\u003Cnumber>` where `\u003Cnumber>` increments only when the package has changes\n  - `released`: `X.Y.Z` using semver per package\n- For alpha\u002Fbeta date stamps, use the current Pacific date as the cutoff, not UTC and not the user's local\n  timezone. Same-Pacific-day re-cuts use a `.postN` suffix. Honor an explicit user-provided date over this\n  default.\n- Keep the `Development Status` classifier in `pyproject.toml` aligned with the lifecycle stage:\n  - `alpha` -> `Development Status :: 3 - Alpha`\n  - `beta` -> `Development Status :: 4 - Beta`\n  - `rc` -> `Development Status :: 4 - Beta`\n  - `released` -> `Development Status :: 5 - Production\u002FStable`\n- See the PyPI classifier list for the available classifier values:\n  `https:\u002F\u002Fpypi.org\u002Fclassifiers\u002F`\n\n## Installation Options\n\n```bash\npip install agent-framework-core          # Core only\npip install agent-framework-core[all]     # Core + curated standard integrations\npip install agent-framework               # Same standard install set as core[all]\npip install agent-framework-foundry       # Specific connector (pulls in core)\n```\n\n## Maintaining Documentation\n\nWhen changing a package, check if its `AGENTS.md` needs updates:\n- Adding\u002Fremoving\u002Frenaming public classes or functions\n- Changing the package's purpose or architecture\n- Modifying import paths or usage patterns\n\nKeep `python\u002FPACKAGE_STATUS.md` updated when:\n- A package is added, removed, renamed, or promoted between lifecycle stages\n- A package is added to or removed from the root meta-package\n- A package starts or stops exposing individually staged experimental or release-candidate APIs\n\nWhen a package adds, removes, or renames environment variables, update the related documentation in the same\nchange:\n- The package's `README.md` for package-level configuration\u002Fenv var guidance\n- `samples\u002FREADME.md` if the package is included in `packages\u002Fcore\u002Fpyproject.toml` `[all]` and the env var is\n  part of the consolidated package env-var inventory\n- Any affected sample\u002Fpackage-local `.env.example`, `.env.template`, or sample README files when sample setup\n  changes alongside the package\n",{"name":5,"description":7},"main","b7884ad88eb85c27b2a7bf40c0e15948e7ccc28d","python\u002F.github\u002Fskills\u002Fpython-package-management",2624,"https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fagent-framework","https:\u002F\u002Faka.ms\u002Fagent-framework","https:\u002F\u002Fgithub.com\u002Fmicrosoft","https:\u002F\u002Fopensource.microsoft.com","Open source projects and samples from Microsoft","2026-11-14 14:06:55",{"source":8,"skillName":5,"commands":54,"prompt":64,"zipAvailable":65},[55,58,61],{"tool":56,"command":57},"npx","npx skills add microsoft\u002Fagent-framework --skill python-package-management",{"tool":59,"command":60},"bunx","bunx skills add microsoft\u002Fagent-framework --skill python-package-management",{"tool":62,"command":63},"pnpm","pnpm dlx skills add microsoft\u002Fagent-framework --skill python-package-management","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fagent-framework\u002Fblob\u002Fmain\u002Fpython\u002F.github\u002Fskills\u002Fpython-package-management\u002FSKILL.md",true,[67],{"path":68,"resourceType":69,"resourceTypeDesc":70,"sizeBytes":71,"sourceHtmlUrl":72},"agents\u002Fopenai.yaml","REFERENCE","参考文档",275,"https:\u002F\u002Fgithub.com\u002Fshengdabai\u002FTony-Claude-Code-Skills\u002Fblob\u002Fmain\u002Fskills\u002F.agents\u002Fskills\u002Fgstack-guard\u002Fagents\u002Fopenai.yaml",[74,97,125,135,158,175],{"skillNo":75,"name":76,"nameZh":6,"description":77,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":78,"categories":79,"keywords":80,"publishedAt":38},"01M70H1GNJPW85ENXA0BCPC6PF","draft-release-notes","Create or rewrite Agent Framework release notes from a generated GitHub release log. Use when preparing Python or .NET release notes, filtering mixed-language changelogs, deduplicating changes, preserving inline contributor credits, or documenting breaking-change migrations from pull requests. Requires the user to provide the generated GitHub log, language, and release version before work begins.","https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fagent-framework\u002Fblob\u002Fmain\u002F.github\u002Fskills\u002Fdraft-release-notes\u002FSKILL.md",[],[81,82,83,84,85,19,37,86,87,88,89,90,31,91,92,93,94,95,96,24],"notes","release notes","release","draft","draft release notes","pull requests","create","pull","generated","filtering","version","log","release version","rewrite","mixed-language","contributor",{"skillNo":98,"name":99,"nameZh":6,"description":100,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":101,"categories":102,"keywords":108,"publishedAt":38},"01M699C55AYGT1ZVZDHWFZNBDB","agent-framework-builder-python","Use this skill when building, modifying, reviewing, evaluating, integrating, or hosting Python applications with Microsoft Agent Framework. Use it for Agent Framework agents, harness agents, workflows, orchestrations, declarative definitions, tools, sessions, middleware, evaluation, observability, integrations, or hosting, even when the user has not yet chosen the correct framework pattern.","https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fagent-framework\u002Fblob\u002Fmain\u002Fskills\u002Fagent-framework-builder-python\u002FSKILL.md",[103],{"categoryNo":104,"categoryName":105,"taxonomyType":106,"taxonomyTypeDesc":107},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[19,37,109,35,110,9,111,112,113,114,115,116,117,118,119,31,120,29,121,122,123,27,124,24],"agent framework builder python","builder","workflows","sessions","microsoft agent framework","harness","observability","Evaluation","applications","correct","integrations","hosting","middleware","with","orchestrations","declarative",{"skillNo":126,"name":127,"nameZh":6,"description":128,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":129,"categories":130,"keywords":132,"publishedAt":38},"01M699C2PH4WPRCMV90NHANCY9","agent-framework-builder-dotnet","Use this skill when building, modifying, reviewing, evaluating, integrating, or hosting .NET (C#) applications with Microsoft Agent Framework. Use it for Agent Framework agents, harness agents, workflows, orchestrations, declarative definitions, tools, sessions, middleware, evaluation, observability, integrations, or hosting, even when the user has not yet chosen the correct framework pattern.","https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fagent-framework\u002Fblob\u002Fmain\u002Fskills\u002Fagent-framework-builder-dotnet\u002FSKILL.md",[131],{"categoryNo":104,"categoryName":105,"taxonomyType":106,"taxonomyTypeDesc":107},[37,133,134,35,110,9,111,112,113,114,115,117,118,119,31,121,122,123,27,124,24],"dotnet","agent framework builder dotnet",{"skillNo":136,"name":137,"nameZh":6,"description":138,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":139,"categories":140,"keywords":142,"publishedAt":38},"01M699C0QKV5XEBE61PB8NV544","foundry-hosted-agent-validation","Step-by-step process for validating a Python Foundry hosted agent sample (under python\u002Fsamples\u002F04-hosting\u002Ffoundry-hosted-agents\u002F) end to end — running it locally (native runtime and `azd ai agent run`) and after deploying it to an Azure AI Foundry project with `azd`. Use this when asked to validate a hosted agent sample.","https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fagent-framework\u002Fblob\u002Fmain\u002Fpython\u002Fscripts\u002Fsample_validation\u002Fskills\u002Ffoundry-hosted-agent-validation\u002FSKILL.md",[141],{"categoryNo":104,"categoryName":105,"taxonomyType":106,"taxonomyTypeDesc":107},[143,144,145,19,146,147,148,149,35,150,151,31,152,153,154,155,156,27,24,157],"foundry","foundry hosted agent validation","validation","ai agent","azure","run","azure ai foundry","tool","model","project","sample","runtime","end","native","running",{"skillNo":159,"name":160,"nameZh":6,"description":161,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":162,"categories":163,"keywords":164,"publishedAt":38},"01M699BYCYJD8RXC5MYY61WBPR","foundry-config-setup","Resolve missing setup caused by a hardcoded Foundry project endpoint or model in a sample. Use when a sample fails because it uses a placeholder\u002Fhardcoded project_endpoint (for example \"https:\u002F\u002Fyour-project.services.ai.azure.com\") or a hardcoded model instead of reading them from the environment.","https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fagent-framework\u002Fblob\u002Fmain\u002Fpython\u002Fscripts\u002Fsample_validation\u002Fskills\u002Ffoundry-config-setup\u002FSKILL.md",[],[143,165,166,167,168,147,169,170,171,150,151,172,31,173,152,153,174,122,27,24],"foundry config setup","config","setup","placeholder","tool use","https","services","environment","example","missing",{"skillNo":176,"name":177,"nameZh":6,"description":178,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":179,"categories":180,"keywords":181,"publishedAt":38},"01M699BWSTE6SGCCBY3JFP10GQ","support-style","Contoso Outdoors customer-support tone and formatting guidelines.","https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fagent-framework\u002Fblob\u002Fmain\u002Fpython\u002Fsamples\u002F04-hosting\u002Ffoundry-hosted-agents\u002Fresponses\u002Ffoundry_toolbox_mcp_skills\u002Fskills\u002Fsupport-style\u002FSKILL.md",[],[182,183,184,185,186,187,188],"support","support style","style","customer","guidelines","tone","formatting"]