microsoft/power-platform-skills is a plugin marketplace containing reusable skills, agents, and commands for developing with Microsoft Power Platform. Developers use it to build and deploy Power Pages sites, model-driven Power Apps, and related solutions through Claude Code or GitHub Copilot. The catalogue entries are the marketplace's included skills, agents, plugins, and other agent components.
Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/microsoft/power-platform-skillsnpx agentmods add skills/microsoft/power-platform-skills/plan-almWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/microsoft/power-platform-skills/plan-alm)<a href="https://agentmods.dev/skills/microsoft/power-platform-skills/plan-alm"><img src="https://agentmods.dev/badge/skills/microsoft/power-platform-skills/plan-alm/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/microsoft/power-platform-skills/plan-alm"><img src="https://agentmods.dev/badge/skills/microsoft/power-platform-skills/plan-alm.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 12 findings, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high YARA Match · line 3 YARA rule matched a hack tool or exploit indicator (offensive tools, reconnaissance, privilege escalation, or exploit frameworks).Fix: Remove offensive tool references and exploit code. Legitimate agent skills should not contain penetration testing tools, exploit frameworks, or reconnaissance utilities.
- high Prompt Injection · line 51 Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.Fix: Audit all comments and invisible characters. Remove any instructions that direct the agent to perform unauthorized actions. Use plain, reviewable content.
- high Prompt Injection · line 318 Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.Fix: Audit all comments and invisible characters. Remove any instructions that direct the agent to perform unauthorized actions. Use plain, reviewable content.
- high Privilege Escalation · line 339 Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.Fix: Remove references to credential paths. Use environment variables or secrets managers. For docs, use placeholder paths (e.g., /path/to/config). Never load .env or token files in production code paths.
- high Prompt Injection · line 408 Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.Fix: Audit all comments and invisible characters. Remove any instructions that direct the agent to perform unauthorized actions. Use plain, reviewable content.
- high Prompt Injection · line 509 Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.Fix: Audit all comments and invisible characters. Remove any instructions that direct the agent to perform unauthorized actions. Use plain, reviewable content.
- high Prompt Injection · line 530 Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.Fix: Audit all comments and invisible characters. Remove any instructions that direct the agent to perform unauthorized actions. Use plain, reviewable content.
- high Prompt Injection · line 594 Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.Fix: Audit all comments and invisible characters. Remove any instructions that direct the agent to perform unauthorized actions. Use plain, reviewable content.
- high Prompt Injection · line 620 Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.Fix: Audit all comments and invisible characters. Remove any instructions that direct the agent to perform unauthorized actions. Use plain, reviewable content.
- medium Excessive Agency · line 18 Skill selects an external model or provider that may use a different account or billing plan than the operator expects. Undisclosed model switches can cause unexpected cost or quota consumption.Fix: Remove the model/provider override or disclose it prominently and require explicit operator approval before invoking an external coding CLI or billed model.
- medium Excessive Agency · line 574 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
- medium Excessive Agency · line 575 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00185 | $0.24888 |
| Opus 5.5 | $0.00074 | $0.09955 |
| Sonnet 5.5 | $0.00037 | $0.04978 |
| Haiku 4.5 | $0.00018 | $0.02489 |
Grade A, and why
plan-alm scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Runs shell commandslowCapability
Expected in a hook, worth knowing in a rule or an instructions file.
> **Why no Node wrapper.** The earlier `node -e "spawn('powershell.exe', [...])"` chain hits the agent's sandbox classifier (Node spawning a process that spawns another process is the textbook pattern the classifier bloc How it starts
The opening of the file, as written. The whole thing — 1,134 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Plugin check: Run
node "${PLUGIN_ROOT}/scripts/check-version.js"— if it outputs a message, show it to the user before proceeding.
plan-alm
A 4-phase planner that gathers ALM strategy from the user, generates an HTML deployment plan, and gets approval. It does not execute anything — execution is delegated to the individual ALM skills, which the user runs afterward.
Overview
This skill detects the current project state (existing solution, pipeline), asks targeted questions about the desired promotion strategy (Power Platform Pipelines or Manual export/import), generates a visual docs/alm-plan.html, and gets user approval. The four phases are: Phase 1 — Detect, Phase 2 — Gather strategy, Phase 3 — Generate plan, Phase 4 — Approve & save.
plan-alm never deploys. The plan's steps[] array records the recommended execution sequence. After approval, the user invokes the individual skills — setup-solution, setup-pipeline (or export-solution), and deploy-pipeline (or import-solution) — in that order. Each of those skills detects the approved plan via its Phase 0 gate, proceeds without re-nagging, and refreshes the plan on completion. This separation is deliberate: it keeps plan-alm safe to run unattended (e.g. under autopilot) because no single answer can trigger an irreversible deployment.
Do NOT create tasks at the start — strategy is unknown until Phase 2 completes. Create both tasks in Phase 3 once the strategy is determined.
Phase 1 — Detect Project State
Do NOT create tasks yet. Use natural language progress reporting only during this phase.
Steps:
- Detect prior ALM deferral for this project. Before any discovery work, check whether the project root contains a
.alm-deferredmarker file. The marker is written by users who explicitly opted ALM-skill validators out of "missing artifacts" warnings (e.g. "this site is handled separately" or "ni-dev — no ALM"). If a user is now invokingplan-alm, we should surface that the marker is present and ask what to do, rather than silently proceeding (which would build a plan the user previously decided not to maintain) or silently removing the marker (which would re-enable nags on every other ALM skill).
What ships with it
3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday Changed e412da7d19ad
- 1mo ago First seen · 1,134 lines · 185 tokens per session scan A f2e479821541
plan-alm is a skill published in the GitHub repository microsoft/power-platform-skills (1,061 stars, last pushed yesterday), licensed MIT. It adds 185 tokens to every session and 24,888 once invoked, about $0.0007 per session on Opus 5.5. A static security scan graded it A with 1 finding (runs shell commands). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other skills, from other repositories
canvas-apps-ui-gen
Generates paste-ready Power Apps Canvas App YAML. Invoke when the user wants to replicate a UI mockup, improve an existing Canvas app screen, or build a new screen from a text description. Also invoke when the user asks to "improve", "redesign", or "generate YAML" for a Canvas app screen.
top-design
Create award-winning, immersive web experiences at the level of Awwwards-featured agencies. Use when the user mentions "Awwwards quality", "make my site stunning", "scroll animations", "parallax storytelling", "cinematic web design", "portfolio site", or "brand experience". Also trigger when elevating a standard…
ux-heuristics
Evaluate and improve interface usability using heuristic analysis. Use when the user mentions "usability audit", "users are confused", "form usability", "navigation problems", "Nielsen heuristics", "cognitive walkthrough", or "is this easy to use". Also trigger when reviewing a design for usability issues, improving…
blue-ocean-strategy
Create uncontested market space using value innovation instead of competing head-to-head. Use when the user mentions "blue ocean", "red ocean", "strategy canvas", "ERRC framework", "value innovation", "non-customers", "buyer utility map", "the market is too crowded", "how do we stand out", or "escape the price war".…
clean-code
Write readable, maintainable code through disciplined naming, small functions, and clean error handling. Use when the user mentions "clean up this code", "this function is too long", "code smells", "naming conventions", "boy scout rule", "single responsibility", or "unit test quality". Also trigger when reviewing a…
session-scan
Compute metrics for Claude Code sessions. Discovers via ccrider, filters trivial, computes friction/opportunity/fingerprint scores. Use for broad session triage.