HVE Core is a collection of agents, prompts, coding instructions, and skills for building repeatable software-development workflows with GitHub Copilot. It is intended for individuals and teams that want structured AI-assisted research, planning, implementation, and review, while the catalogue entries provide many of its reusable workflow components.
Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add microsoft/hve-core --skill hve-core-installergit clone --depth 1 https://github.com/microsoft/hve-coreWrote 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/hve-core/hve-core-installer)<a href="https://agentmods.dev/skills/microsoft/hve-core/hve-core-installer"><img src="https://agentmods.dev/badge/skills/microsoft/hve-core/hve-core-installer/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/hve-core/hve-core-installer"><img src="https://agentmods.dev/badge/skills/microsoft/hve-core/hve-core-installer.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 9 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 Prompt Injection · line 80 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 152 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 226 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 269 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 307 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 329 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 413 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 Tool Misuse · line 445 Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).Fix: Validate all tool parameters against an allowlist. Reject dangerous parameter values (shell=True, --force, -rf /) and use safe defaults.
- high Tool Misuse · line 449 Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).Fix: Validate all tool parameters against an allowlist. Reject dangerous parameter values (shell=True, --force, -rf /) and use safe defaults.
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 | $0.00055 | $0.06542 |
| Opus 5.5 | $0.00022 | $0.02617 |
| Sonnet 5.5 | $0.00011 | $0.01308 |
| Haiku 4.5 | $0.00006 | $0.00654 |
Grade C, and why
hve-core-installer scanned grade C 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.
Recursive force deletehighDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
| 1 (Peer Clone) | `rm -rf ../hve-core` | How it starts
The opening of the file, as written. The whole thing — 610 lines — stays where its author put it; the contents beside it link to each section on GitHub.
HVE-Core Installer Skill
Decision-driven installer for HVE-Core in VS Code's Copilot agent harness. It recommends the HVE-Core plugin first, then a team default in .github/copilot/settings.json, then native repository adoption, which copies every HVE-Core component into the project's own .github folders.
The VS Code extension remains an editor option. A clone is a storage choice that Copilot uses as a plugin from source or, for a copy inside the project's checkout, through folders declared in .github/copilot/settings.json. The skill also validates the result and configures MCP servers.
Role Definition
Operate as two collaborating personas:
- The Installer persona detects the environment, guides route selection, and executes installation steps
- The Validator persona verifies installed folders and repository settings, then has the user confirm that HVE-Core agents and skills appear in Copilot Chat
The Installer persona handles all detection and execution. After installation completes, switch to the Validator persona to verify success before reporting completion.
Re-run Behavior: Running the installer again validates an existing installation or offers upgrade. Safe to re-run anytime.
Required Phases
| Phase | Name | Purpose |
|---|---|---|
| 1 | Environment Detection | Obtain consent and check for earlier HVE-Core installs |
| 2 | Install Route Selection | Choose the plugin, team default, native adoption, extension, or clone route |
| 3 | Environment Detection & Decision Matrix | For native adoption and clones: detect the environment and recommend where to keep the clone |
| 4 | Installation Methods | Clone HVE-Core and, for the clone route, connect the clone to Copilot |
| 5 | Validation | Verify the clone before it is used |
| 6 | Post-Installation Setup | Configure gitignore and MCP servers, then reload the window and check Copilot Chat |
| 7 | Component Installation | Native adoption: copy every component into the project's .github folders |
What ships with it
29 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.
- examples/README.md 3.9 KB
- references/component-installation.md 16 KB
- references/installation-methods.md 15 KB
- references/post-installation-setup.md 12 KB
- scripts/collision-detection.ps1 1.6 KB runs code
- scripts/collision-detection.sh 1006 B runs code
- scripts/component-copy.ps1 20 KB runs code
- scripts/component-copy.sh 17 KB runs code
- scripts/detect-environment.ps1 1.5 KB runs code
- scripts/detect-environment.sh 1.3 KB runs code
- scripts/eject.ps1 3.8 KB runs code
- scripts/eject.sh 2.5 KB runs code
- scripts/file-status-check.ps1 2.6 KB runs code
- scripts/file-status-check.sh 2.2 KB runs code
- scripts/upgrade-detection.ps1 2.5 KB runs code
- scripts/upgrade-detection.sh 2.1 KB runs code
- scripts/validate-extension.ps1 1.3 KB runs code
- scripts/validate-extension.sh 1.1 KB runs code
- scripts/validate-installation.ps1 2.1 KB runs code
- scripts/validate-installation.sh 1.8 KB runs code
- SECURITY.md 30 KB
- tests/collision-detection.Tests.ps1 8.1 KB runs code
- tests/component-copy.Tests.ps1 60 KB runs code
- tests/detect-environment.Tests.ps1 3.8 KB runs code
- tests/eject.Tests.ps1 15 KB runs code
- tests/file-status-check.Tests.ps1 8.4 KB runs code
- tests/upgrade-detection.Tests.ps1 9.9 KB runs code
- tests/validate-extension.Tests.ps1 4.3 KB runs code
- tests/validate-installation.Tests.ps1 8.9 KB runs code
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 · +119 lines · +27 tokens per session a6ebd05e19fe
- 1mo ago Changed · -1 lines 5d454295a466
- 2mo ago Changed 8b15bb518616
- 2mo ago First seen · 492 lines · 28 tokens per session scan C 0c0bd688144d
hve-core-installer is a skill published in the GitHub repository microsoft/hve-core (1,606 stars, last pushed today), licensed MIT. It adds 55 tokens to every session and 6,542 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it C with 1 finding (recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
hve-core-installer compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| hve-core-installermicrosoft/hve-core | Skill | 1.6k | today | MIT | C | 55 | Claude Code |
| imagegenopenai/codex | Skill | 134k | today | Apache-2.0 | A | 113 | Codex |
| local-ai-agentsmicrosoft/ai-agents-for-beginners | Skill | 79k | yesterday | MIT | A | 200 | Claude Code, Codex |
| playwright-climicrosoft/playwright | Skill | 100k | yesterday | Apache-2.0 | A | 19 | Claude Code |
| browser-automationopenclaw/openclaw | Skill | 394k | today | MIT | A | 36 | Claude Code |
| systematic-debuggingobra/superpowers | Skill | 314k | 4d ago | MIT | A | 21 | Claude Code |
| next-cache-components-adoptionvercel/next.js | Skill | 143k | today | MIT | A | 95 | Claude Code |