You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

47 lines
1.9 KiB

Ai/docs alignment (#91) * chore: initialize the memory bank Add the canonical .memory-bank base with evidence-based project context: purpose and scope, workflows, stack and validation commands, architecture map, decisions, and the open work found while documenting the cmdlets. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: AI Assistant <ai@example.com> * docs: align documentation with the cmdlet source - Fill all 36 platyPS cmdlet pages from the C# source: synopsis, description, parameters, defaults, examples, inputs, outputs, and notes, including documented limitations of the current code - Check every example against live parameter metadata and run them in a sandbox; fix examples that did not work (CSV restore, account filter, recursive inheritance, -AccessRights typos) - Rewrite the home, concepts, examples, README, and contributor pages; add a grouped cmdlet overview, module settings, privileges, long paths, and the platyPS workflow - Document Remove-Item2 -PassThru as renamed after 4.2.6 (#64) - Fix mkdocs.yml navigation, edit_uri, and copyright markup; add build.os and a pinned MkDocs version for Read the Docs - Point online help links to the pages on GitHub; add CHANGELOG.md Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: AI Assistant <ai@example.com> * ci: build the module and check the docs against the build The documentation check ran Update-MarkdownHelp against the NTFSSecurity release from the PowerShell Gallery (4.2.6), so it failed for every unreleased parameter change. PR #91 failed because 4.2.6 still has Remove-Item2 -PassThur while the source and the docs have -PassThru. - Build NTFSSecurity.csproj in Release on the Visual Studio 2022 image, using the .NET Framework 4.5.2 reference assemblies package instead of an installed targeting pack - Check Docs/Cmdlets against the module built from source - Pin platyPS 0.14.2 and MarkdownLinkCheck 0.2.0, and enable TLS 1.2 so the NuGet provider bootstrap works Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: AI Assistant <ai@example.com> * chore: record the green PR 91 build in the memory bank Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: AI Assistant <ai@example.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Co-authored-by: AI Assistant <ai@example.com>
4 weeks ago
---
schema-version: 1
loading-mode: routed
status: accepted
owner: shared
last-verified: 2026-10-02
source: repository evidence
---
# Memory Bank index
Read this file first. It routes tasks to the smallest relevant set of Memory
Bank files.
## Full-read fallback
Set `loading-mode` to `full` to restore complete-base loading. Also fail open
when this index is missing or invalid, the task is ambiguous, routes conflict,
a listed file is missing, or a critical fact cannot be found. Full mode also
reads every existing `decisions/*.md` record.
## Routing table
Combine routes when a task spans topics. For durable repository writes, also
read `activeContext.md` before editing.
| Route | Task signals | Read |
|---|---|---|
| `general` | General Q&A with no project decision | Index only |
| `continuation` | Resume, current focus, next step | `activeContext.md`, `progress.md` |
| `scope` | Purpose, scope, requirements, Acceptance criteria | `projectbrief.md` |
| `product` | Users, problem, workflow, experience goal | `productContext.md` |
| `implementation` | Code, configuration, build, test, dependency, deployment | `techContext.md`, `activeContext.md` |
| `architecture` | Design, pattern, decision, migration, integration | `systemPatterns.md`, relevant `decisions/*.md` |
| `status` | Progress, recent change, open work | `progress.md`, `activeContext.md` |
| `language` | Canonical terms in authored artifacts | `glossary.md` |
| `interaction-history` | Session analysis, prompt trends, Memory Bank evals | `promptHistory.md`, `progress.md` |
| `role` | Active Custom agent domain workflow | Only that agent's declared role files |
## Authority order
1. The user's current request controls task constraints.
2. Repository source, configuration, tests, and evidence control facts.
3. Accepted decision records control durable architectural choices.
4. Core Memory Bank files control only their assigned topic.
5. Historical logs never override current source.