[codex] Add initial skills library

.editorconfig

@@ -0,0 +1,18 @@
+root = true
+
+[*]
+charset = utf-8
+end_of_line = lf
+insert_final_newline = true
+indent_style = space
+indent_size = 4
+
+[*.md]
+indent_size = 2
+trim_trailing_whitespace = false
+
+[*.json]
+indent_size = 2
+
+[*.yml]
+indent_size = 2

.gitattributes

@@ -0,0 +1,4 @@
+* text=auto eol=lf
+*.ps1 text eol=lf
+*.psm1 text eol=lf
+*.psd1 text eol=lf

.github/workflows/validate.yml

@@ -0,0 +1,47 @@
+name: validate
+
+on:
+  pull_request:
+  push:
+    branches:
+      - main
+
+permissions:
+  contents: read
+
+jobs:
+  powershell:
+    runs-on: windows-latest
+    steps:
+      - uses: actions/checkout@v7
+      - name: Parse PowerShell scripts
+        shell: pwsh
+        run: |
+          $failed = $false
+          Get-ChildItem -Recurse -Filter *.ps1 | ForEach-Object {
+            $tokens = $null
+            $errors = $null
+            [System.Management.Automation.Language.Parser]::ParseFile($_.FullName, [ref]$tokens, [ref]$errors) | Out-Null
+            if ($errors.Count -gt 0) {
+              $failed = $true
+              Write-Error "$($_.FullName) has parser errors: $($errors.Message -join '; ')"
+            }
+          }
+          if ($failed) { exit 1 }
+      - name: Check Markdown final newlines
+        shell: pwsh
+        run: |
+          $failed = $false
+          Get-ChildItem -Recurse -File -Filter *.md | ForEach-Object {
+            $bytes = [System.IO.File]::ReadAllBytes($_.FullName)
+            if ($bytes.Length -gt 0 -and $bytes[-1] -ne 10) {
+              $failed = $true
+              Write-Error "$($_.FullName) must end with a newline."
+            }
+          }
+          if ($failed) { exit 1 }
+      - name: Analyze synthetic baseline
+        shell: pwsh
+        run: |
+          New-Item -ItemType Directory -Force -Path .\sysadmin-windows-startup-performance\reports,.\sysadmin-windows-startup-performance\state | Out-Null
+          powershell -NoProfile -ExecutionPolicy Bypass -File .\sysadmin-windows-startup-performance\scripts\analyze-startup-baseline.ps1 -BaselinePath .\sysadmin-windows-startup-performance\examples\synthetic-baseline.json -OutputPath .\sysadmin-windows-startup-performance\reports\synthetic-analysis.md -AnalysisOutputPath .\sysadmin-windows-startup-performance\state\synthetic-analysis.json

.gitignore

@@ -0,0 +1,19 @@
+# Generated local performance data
+**/state/*
+**/reports/*
+**/logs/*
+!**/state/.gitkeep
+!**/reports/.gitkeep
+!**/logs/.gitkeep
+
+# PowerShell and editor noise
+*.tmp
+*.log
+*.bak
+*.ps1xml
+.vscode/
+.idea/
+
+# OS noise
+Thumbs.db
+Desktop.ini

AGENTS.md

@@ -0,0 +1,6 @@
+# Agent Instructions
+
+- When creating or editing Markdown files, always end each `.md` file with a final newline.
+- Keep skills small, composable, and safe by default.
+- Do not commit generated reports, state files, logs, or machine-specific diagnostic captures.
+- Prefer synthetic examples over real local machine data in public artifacts.

CONTRIBUTING.md

@@ -0,0 +1,23 @@
+# Contributing
+
+Contributions should keep skills small, composable, and safe by default.
+
+## Guidelines
+
+- Keep `SKILL.md` concise and move detailed material into directly linked `references/` files.
+- Prefer deterministic scripts for repeated or fragile workflows.
+- Keep generated state, reports, logs, and machine-specific artifacts out of git.
+- Avoid destructive behavior in diagnostic skills. When remediation is needed, make it a separate reviewed workflow.
+- Include synthetic examples rather than real machine captures.
+- Put skill-specific usage, safety, data-handling, and validation details inside the owning skill folder.
+
+## Validation
+
+Before opening a pull request:
+
+- Run your runtime's skill validator for every changed skill.
+- Parse or lint bundled scripts with the language's normal tooling.
+- Run the synthetic examples or validation commands documented by each changed skill.
+- Confirm generated outputs remain ignored by git.
+
+The repository workflow also checks PowerShell parsing, Markdown final newlines, and the current synthetic analysis example.

LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 KarmCraft
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

README.md

@@ -0,0 +1,47 @@
+# Skills Library
+
+A public collection of agent skills that I have found useful enough to package, maintain, and share.
+
+The goal is to keep this library practical: composable, reusable skills that solve real workflow problems for humans and agents using `SKILL.md` compatible runtimes. Each skill should be small enough to understand, safe by default, and easy to combine with other skills.
+
+Each top-level folder is a self-contained, directly installable skill.
+
+## Skills
+
+| Skill | Purpose |
+| --- | --- |
+| [`sysadmin-windows-startup-performance`](sysadmin-windows-startup-performance/) | Diagnose Windows startup performance with read-only baselines and data-driven analysis. |
+
+## Install
+
+Copy the skill folder you want into the skill directory used by your agent runtime.
+
+For Codex on Windows:
+
+```powershell
+$SkillName = "<skill-folder>"
+Copy-Item -Recurse ".\$SkillName" "$env:USERPROFILE\.codex\skills\$SkillName"
+```
+
+For other `SKILL.md` compatible tools, use their documented skill location.
+
+## Repository Layout
+
+```text
+<skill-name>/
+  SKILL.md
+  agents/openai.yaml
+  scripts/
+  tools/
+  references/
+  examples/
+  templates/
+```
+
+Use `tools/` for helper CLIs, adapters, or small utilities that support a skill but are not the primary workflow scripts. Keep primary repeatable workflows in `scripts/`, long-form guidance in `references/`, and avoid vendoring large binaries, secrets, or machine-specific executables.
+
+Not every skill uses every optional folder. Generated reports, state files, and logs are intentionally ignored by git.
+
+## Safety
+
+Skills should be safe by default and explicit about their boundaries. Diagnostic skills should collect evidence before recommending changes, and any remediation workflow should be separately reviewed, reversible, and opt-in.

SECURITY.md

@@ -0,0 +1,13 @@
+# Security Policy
+
+## Reporting
+
+Please report security issues privately through GitHub security advisories if available, or by opening a minimal issue that does not include secrets or exploit details.
+
+## Data Handling
+
+Do not attach real diagnostic captures, generated reports, or local inventory unless you have reviewed and sanitized them first.
+
+Before sharing artifacts, remove hostnames, usernames, domains, command-line arguments, environment paths, service or task arguments, secrets, tokens, and local network identifiers.
+
+Collectors should default to least-sensitive output and document any switches that include more detail in the owning skill's safety reference.

sysadmin-windows-startup-performance/SKILL.md

@@ -0,0 +1,74 @@
+---
+name: sysadmin-windows-startup-performance
+description: Collect and analyze read-only Windows startup performance baselines. Use when Codex is asked to diagnose slow Windows boot or login, collect Diagnostics-Performance event log data, inspect startup apps, services, scheduled tasks, resource pressure, recent boot warnings, or prepare conservative data-driven recommendations without changing system configuration.
+---
+
+# Sysadmin Windows Startup Performance
+
+## Overview
+
+Use this skill for read-only Windows startup diagnostics. It collects sanitized startup baselines, analyzes likely bottlenecks, and produces JSON plus Markdown outputs that other skills can consume.
+
+## Files
+
+- `scripts/collect-startup-baseline.ps1`: elevated read-only collector for Diagnostics-Performance events, startup inventory, service/task signals, resources, and recent boot warnings.
+- `scripts/analyze-startup-baseline.ps1`: analyzer that ranks findings and writes a Markdown report.
+- `references/data-model.md`: baseline and analysis schema notes.
+- `references/safety.md`: operating rules and remediation boundaries.
+- `examples/synthetic-baseline.json`: synthetic fixture for analyzer validation.
+- `tools/`: reserved for optional helper CLIs, adapters, or small utilities that support this skill.
+- `templates/local-performance/`: optional project-local diagnostic hub scaffold.
+
+## Workflow
+
+1. Read `references/safety.md` before collecting data or recommending changes.
+2. Create or use a local hub folder with `scripts/`, `state/`, `reports/`, and `logs/`. A template is available at `templates/local-performance/` when this repository is installed as a project.
+3. Run the collector from an elevated PowerShell session. Do not fall back to non-admin collection because protected boot timing logs can be missed.
+4. Run the analyzer on the latest or specified baseline.
+5. Prefer at least three comparable post-reboot baselines before recommending startup changes.
+6. If the user asks for remediation, produce a dry-run plan with rollback notes first. Do not mutate startup apps, registry entries, scheduled tasks, services, drivers, security tools, or Windows settings without explicit approval.
+
+## Commands
+
+Collect a baseline:
+
+```powershell
+.\scripts\collect-startup-baseline.ps1
+```
+
+Analyze the latest baseline from the same hub:
+
+```powershell
+.\scripts\analyze-startup-baseline.ps1
+```
+
+Analyze a specific baseline:
+
+```powershell
+.\scripts\analyze-startup-baseline.ps1 -BaselinePath .\state\startup-baseline-YYYYMMDD-HHMMSS.json
+```
+
+Include full command-line previews only when explicitly needed:
+
+```powershell
+.\scripts\collect-startup-baseline.ps1 -IncludeCommandLines
+```
+
+## Validation
+
+When modifying this skill, run the analyzer against the synthetic baseline from the repository root:
+
+```powershell
+powershell -NoProfile -ExecutionPolicy Bypass -File .\sysadmin-windows-startup-performance\scripts\analyze-startup-baseline.ps1 -BaselinePath .\sysadmin-windows-startup-performance\examples\synthetic-baseline.json -OutputPath .\sysadmin-windows-startup-performance\reports\synthetic-analysis.md -AnalysisOutputPath .\sysadmin-windows-startup-performance\state\synthetic-analysis.json
+```
+
+The command writes ignored `reports/` and `state/` outputs under the skill folder.
+
+## Output
+
+The collector writes `state/startup-baseline-*.json` unless `-NoWrite` or `-OutputPath` is used. The analyzer writes:
+
+- `reports/startup-baseline-analysis-*.md`
+- `state/startup-analysis-*.json`
+
+See `references/data-model.md` before building downstream tooling against these files.

sysadmin-windows-startup-performance/agents/openai.yaml

@@ -0,0 +1,4 @@
+interface:
+  display_name: "Sysadmin Windows Startup Performance"
+  short_description: "Diagnose Windows startup performance."
+  default_prompt: "For a first-time assessment, guide the user through five reboot cycles. After each boot, wait 3 to 5 minutes, then run $sysadmin-windows-startup-performance to collect one read-only startup baseline. After the fifth collection, run the single-baseline analyzer on the latest collection; when comparison is needed, analyze earlier baseline files one at a time and summarize repeated patterns."