MCP Server Builder
anthropics/skills
Guides the design and implementation of Model Context Protocol servers in TypeScript or Python, from tool naming and error messages to evaluation.
This skill should be used when an AI agent, MCP tool, Revit add-in, or automation must inspect or change a live Revit model and needs to prove its target set, choose bounded read-only context…
Install Revit AI Targeting And Safe Execution
$ npx skills add h30190/HJPLUS_Taiwan_Architect_KB --skill revit-ai-targeting-and-safe-execution -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install h30190/HJPLUS_Taiwan_Architect_KB revit-ai-targeting-and-safe-execution --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/h30190/HJPLUS_Taiwan_Architect_KB.git skills-src && mkdir -p .claude/skills && cp -r skills-src/'raw/設計軟體與工具/BIM工具/Revit/AI目標判定與安全執行/revit-ai-targeting-and-safe-execution' .claude/skills/revit-ai-targeting-and-safe-execution && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "revit-ai-targeting-and-safe-execution" agent skill from https://github.com/h30190/HJPLUS_Taiwan_Architect_KB/tree/main/raw/%E8%A8%AD%E8%A8%88%E8%BB%9F%E9%AB%94%E8%88%87%E5%B7%A5%E5%85%B7/BIM%E5%B7%A5%E5%85%B7/Revit/AI%E7%9B%AE%E6%A8%99%E5%88%A4%E5%AE%9A%E8%88%87%E5%AE%89%E5%85%A8%E5%9F%B7%E8%A1%8C/revit-ai-targeting-and-safe-execution into .claude/skills/revit-ai-targeting-and-safe-execution/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "revit-ai-targeting-and-safe-execution", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/h30190/HJPLUS_Taiwan_Architect_KB/tree/main/raw/%E8%A8%AD%E8%A8%88%E8%BB%9F%E9%AB%94%E8%88%87%E5%B7%A5%E5%85%B7/BIM%E5%B7%A5%E5%85%B7/Revit/AI%E7%9B%AE%E6%A8%99%E5%88%A4%E5%AE%9A%E8%88%87%E5%AE%89%E5%85%A8%E5%9F%B7%E8%A1%8C/revit-ai-targeting-and-safe-executionType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add h30190/HJPLUS_Taiwan_Architect_KB --skill revit-ai-targeting-and-safe-execution -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install h30190/HJPLUS_Taiwan_Architect_KB revit-ai-targeting-and-safe-execution --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/h30190/HJPLUS_Taiwan_Architect_KB.git skills-src && mkdir -p .agents/skills && cp -r skills-src/'raw/設計軟體與工具/BIM工具/Revit/AI目標判定與安全執行/revit-ai-targeting-and-safe-execution' .agents/skills/revit-ai-targeting-and-safe-execution && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "revit-ai-targeting-and-safe-execution" agent skill from https://github.com/h30190/HJPLUS_Taiwan_Architect_KB/tree/main/raw/%E8%A8%AD%E8%A8%88%E8%BB%9F%E9%AB%94%E8%88%87%E5%B7%A5%E5%85%B7/BIM%E5%B7%A5%E5%85%B7/Revit/AI%E7%9B%AE%E6%A8%99%E5%88%A4%E5%AE%9A%E8%88%87%E5%AE%89%E5%85%A8%E5%9F%B7%E8%A1%8C/revit-ai-targeting-and-safe-execution into .agents/skills/revit-ai-targeting-and-safe-execution/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "revit-ai-targeting-and-safe-execution", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add h30190/HJPLUS_Taiwan_Architect_KB --skill revit-ai-targeting-and-safe-execution -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install h30190/HJPLUS_Taiwan_Architect_KB revit-ai-targeting-and-safe-execution --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/h30190/HJPLUS_Taiwan_Architect_KB.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/'raw/設計軟體與工具/BIM工具/Revit/AI目標判定與安全執行/revit-ai-targeting-and-safe-execution' .cursor/skills/revit-ai-targeting-and-safe-execution && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "revit-ai-targeting-and-safe-execution" agent skill from https://github.com/h30190/HJPLUS_Taiwan_Architect_KB/tree/main/raw/%E8%A8%AD%E8%A8%88%E8%BB%9F%E9%AB%94%E8%88%87%E5%B7%A5%E5%85%B7/BIM%E5%B7%A5%E5%85%B7/Revit/AI%E7%9B%AE%E6%A8%99%E5%88%A4%E5%AE%9A%E8%88%87%E5%AE%89%E5%85%A8%E5%9F%B7%E8%A1%8C/revit-ai-targeting-and-safe-execution into .cursor/skills/revit-ai-targeting-and-safe-execution/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "revit-ai-targeting-and-safe-execution", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/h30190/HJPLUS_Taiwan_Architect_KB.git --path 'raw/設計軟體與工具/BIM工具/Revit/AI目標判定與安全執行/revit-ai-targeting-and-safe-execution'--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add h30190/HJPLUS_Taiwan_Architect_KB --skill revit-ai-targeting-and-safe-execution -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install h30190/HJPLUS_Taiwan_Architect_KB revit-ai-targeting-and-safe-execution --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/h30190/HJPLUS_Taiwan_Architect_KB.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/'raw/設計軟體與工具/BIM工具/Revit/AI目標判定與安全執行/revit-ai-targeting-and-safe-execution' .gemini/skills/revit-ai-targeting-and-safe-execution && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "revit-ai-targeting-and-safe-execution" agent skill from https://github.com/h30190/HJPLUS_Taiwan_Architect_KB/tree/main/raw/%E8%A8%AD%E8%A8%88%E8%BB%9F%E9%AB%94%E8%88%87%E5%B7%A5%E5%85%B7/BIM%E5%B7%A5%E5%85%B7/Revit/AI%E7%9B%AE%E6%A8%99%E5%88%A4%E5%AE%9A%E8%88%87%E5%AE%89%E5%85%A8%E5%9F%B7%E8%A1%8C/revit-ai-targeting-and-safe-execution into .gemini/skills/revit-ai-targeting-and-safe-execution/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "revit-ai-targeting-and-safe-execution", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install h30190/HJPLUS_Taiwan_Architect_KB revit-ai-targeting-and-safe-executionInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add h30190/HJPLUS_Taiwan_Architect_KB --skill revit-ai-targeting-and-safe-execution -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/h30190/HJPLUS_Taiwan_Architect_KB.git skills-src && mkdir -p .github/skills && cp -r skills-src/'raw/設計軟體與工具/BIM工具/Revit/AI目標判定與安全執行/revit-ai-targeting-and-safe-execution' .github/skills/revit-ai-targeting-and-safe-execution && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "revit-ai-targeting-and-safe-execution" agent skill from https://github.com/h30190/HJPLUS_Taiwan_Architect_KB/tree/main/raw/%E8%A8%AD%E8%A8%88%E8%BB%9F%E9%AB%94%E8%88%87%E5%B7%A5%E5%85%B7/BIM%E5%B7%A5%E5%85%B7/Revit/AI%E7%9B%AE%E6%A8%99%E5%88%A4%E5%AE%9A%E8%88%87%E5%AE%89%E5%85%A8%E5%9F%B7%E8%A1%8C/revit-ai-targeting-and-safe-execution into .github/skills/revit-ai-targeting-and-safe-execution/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "revit-ai-targeting-and-safe-execution", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add h30190/HJPLUS_Taiwan_Architect_KB --skill revit-ai-targeting-and-safe-execution -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install h30190/HJPLUS_Taiwan_Architect_KB revit-ai-targeting-and-safe-execution --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/h30190/HJPLUS_Taiwan_Architect_KB.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/'raw/設計軟體與工具/BIM工具/Revit/AI目標判定與安全執行/revit-ai-targeting-and-safe-execution' .opencode/skills/revit-ai-targeting-and-safe-execution && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "revit-ai-targeting-and-safe-execution" agent skill from https://github.com/h30190/HJPLUS_Taiwan_Architect_KB/tree/main/raw/%E8%A8%AD%E8%A8%88%E8%BB%9F%E9%AB%94%E8%88%87%E5%B7%A5%E5%85%B7/BIM%E5%B7%A5%E5%85%B7/Revit/AI%E7%9B%AE%E6%A8%99%E5%88%A4%E5%AE%9A%E8%88%87%E5%AE%89%E5%85%A8%E5%9F%B7%E8%A1%8C/revit-ai-targeting-and-safe-execution into .opencode/skills/revit-ai-targeting-and-safe-execution/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "revit-ai-targeting-and-safe-execution", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
Facts
- Skill name
revit-ai-targeting-and-safe-execution- GitHub stars
- 158
- Token cost
- ~3.6k tokens
- SKILL.md length
- 1,525 words
- Files
- 2 (incl. references)
- Skills in repo
- 86
- Repo updated
- First seen
- Licence
- CC-BY-SA-4.0
At a glance
This skill should be used when an AI agent, MCP tool, Revit add-in, or automation must inspect or change a live Revit model and needs to prove its target set, choose bounded read-only context…
- Works in 7 steps: Read Current Revit State → Prove the Target Set → Inspect Read-Only Context Progressively → …
- Tasks that involve MCP servers
- SKILL.md covers Overview, Non-Negotiable Distinctions, Execution Steps and Worked Example, plus 5 more sections
- Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md
What it does
Revit AI Targeting And Safe Execution is an agent skill from h30190/HJPLUS_Taiwan_Architect_KB. This skill should be used when an AI agent, MCP tool, Revit add-in, or automation must inspect or change a live Revit model and needs to prove its target set, choose bounded read-only context, preview risk, execute in a supported Revit API context, read the result back independently, and report evidence without overstating success.
Its SKILL.md is about 3.6k tokens, which your agent loads only when the skill is triggered. The skill folder holds 2 other files, including reference files (for example `references/evidence-contract.md`). Compatibility notes: claude-code,opencode,agent-skills
It sits in Agent Workflows, covering MCP servers. The repository describes itself as: 一個開源的台灣AEC產業知識庫，只要你願意共享經驗與知識就歡迎加入我們. The licence is CC-BY-SA-4.0.
When your agent uses it
- Tasks that involve MCP servers
Example prompts
- “/revit-ai-targeting-and-safe-execution”
Requirements
- Compatibility (from SKILL.md): claude-code,opencode,agent-skills
Workflow steps
7 steps, taken from the step headings in SKILL.md.
What it can do on your machine
Read from SKILL.md and the folder at commit 35ed06e. It shows what the files ask for, not the result of running them.
Tool permissions
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Runs code
No scripts in the folder and no shell commands in SKILL.md.
From the folder's file list and the shell code blocks in SKILL.md.
Network
Links to these hosts (documentation or services it may open):
help.autodesk.comgithub.com
From URLs in SKILL.md, links to its own repository left out.
Credentials
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Compatibility
claude-code,opencode,agent-skills
From compatibility in the SKILL.md frontmatter.
Context cost
Revit AI Targeting And Safe Execution loads about 3.6k tokens when it runs, and up to ~4.5k if it reads all its reference files. Until then it costs about 93 tokens; SKILL.md has 1,525 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
Safety
Auto-check passed
The automated check found no risky patterns in SKILL.md.
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.
SKILL.md
The full file from h30190/HJPLUS_Taiwan_Architect_KB at commit 35ed06e, republished under its CC-BY-SA-4.0 licence (© h30190). 1,525 words, ~3,605 tokens.
Download SKILL.mdSave it as
.claude/skills/revit-ai-targeting-and-safe-execution/SKILL.md (or your agent's skills folder). This skill also uses 1 other file; get the full folder from GitHub.- name
- revit-ai-targeting-and-safe-execution
- description
- This skill should be used when an AI agent, MCP tool, Revit add-in, or automation must inspect or change a live Revit model and needs to prove its target set, choose bounded read-only context, preview risk, execute in a supported Revit API context, read the result back independently, and report evidence without overstating success.
- compatibility
- claude-code,opencode,agent-skills
- type
- Skill
- license
- CC-BY-SA-4.0
- metadata.audience
- architects
- metadata.region
- taiwan
- metadata.class
- A
- metadata.data-currency
- 2026-08-28
Revit AI Targeting and Safe Execution
Overview
Use this skill for any live Revit task where an AI-controlled workflow may inspect, create, modify, or delete model data. The workflow is:
read current state → prove targets → inspect only what is needed → preview risk and scope → execute through Revit → read back independently → report evidence and recovery
This is a workflow skill, not an installation guide, product API, or claim that one tool makes arbitrary Revit automation safe. Autodesk documentation is authoritative for Revit API context and transaction behavior. BIM Personal Agent is the contributor's personal research project; v0.8.1 is cited only as a public implementation example of read-only element inspection, explicit target sources, and conservative single-target auto-follow.[^bim-personal-agent-v081]
No existing Revit skill in this knowledge base covers this workflow. The cross-linked uncertainty skill governs source confidence; this skill governs live-model targeting, execution, and readback.
Non-Negotiable Distinctions
Keep these claims separate:
- Target resolved: the exact element set is known.
- Execution returned: Revit or a bridge returned a result.
- Model impact observed: created, modified, or deleted IDs are known.
- Outcome verified: required claims were independently read from the model.
- Deployment verified: the intended manifest, DLL, runtime, and configuration are installed.
None of these proves the others.
Execution Steps
1. Read Current Revit State
Before planning a mutation, capture only the state needed to bind the task:
- exact Revit version;
- connection or bridge identity, when one exists;
- active document identity or a privacy-safe fingerprint;
- active view ID and type when view context matters;
- current selection IDs;
- whether the document changed during inspection.
If the active document changes between planning and execution, invalidate the plan and re-read state.
2. Prove the Target Set
Classify the target source before using it:
| Target source | May execute automatically? | Required check |
|---|---|---|
| Explicit ElementId supplied by the user | Yes | Confirm it exists in the bound document and matches the requested kind |
| Current selection | Yes, when selection and intent match | Preserve the complete selected ID set; do not silently take only the first item |
| Reviewed candidate set | Yes | Record the reviewed IDs and the distinguishing evidence |
| IDs returned by an earlier tool in the same trace | Yes | Preserve lineage to the producing request |
| IDs created by the current operation | Yes for follow-up/readback | Preserve creation lineage and transaction state |
| Name, category, family, type, proximity, or heuristic match | No | Treat as candidates and request review or add a deterministic filter |
An empty target or ambiguous candidate set is a stop condition, not permission to guess.
Single-target auto-follow is a conservative convenience: automatically deepen inspection only when exactly one target is proven. It is not a ban on multi-target work. A multi-target mutation is allowed when the full set is explicit, previewable, and verifiable.
3. Inspect Read-Only Context Progressively
Start with the smallest useful level:
- Summary: identity, category/type, document/view relevance, location summary.
- Parameters: requested instance and type parameters, including writable/read-only state.
- Full: geometry summary, relationships, owner/view/sheet context, or other bounded detail required by the task.
Do not begin with a project-wide scan when one exact element answers the question. Read-only inspection must not open a mutation Transaction or trigger another inspection recursively.
4. Classify Risk and Preview Scope
| Risk class | Minimum preview | Execution rule |
|---|---|---|
| Read-only | Query purpose and target set | No model mutation |
| Reversible write | Target IDs, before/after fields, expected count, Undo label | Execute in a named Transaction and verify Commit status |
| Multi-step reversible write | All steps, shared document, failure behavior | Use an atomic boundary when supported; rollback the group if a required step fails |
| Destructive or hard-to-recover | Actual affected scope and recovery limits | Require explicit human confirmation; refuse if actual scope cannot be shown |
| Uncertain after dispatch | Last confirmed state and unresolved operation | Do not retry automatically; read current state first |
Any Revit model change requires an active Transaction.[^autodesk-revit-2024-transactions] A committed named Transaction appears in Revit's Undo menu. TransactionGroup can rollback committed inner transactions or assimilate them into one Undo item when the API workflow supports it.[^autodesk-revit-2024-transaction-classes]
5. Execute Through a Supported Revit API Context
For a modeless UI or asynchronous agent, marshal work through ExternalEvent or another supported Revit API entry point. Raising an ExternalEvent requests execution; Revit calls the handler when it can process the event.[^autodesk-revit-2024-external-events]
During execution:
- bind again to the expected document;
- restrict inputs to the proven target set;
- give the Transaction a task-specific name;
- record Start, Commit, Rollback, or Pending status;
- collect actual created, modified, and deleted ElementIds;
- never treat queue acceptance as completed execution.
6. Read Back Required Claims Independently
Define the claims before execution. Examples:
- parameter X on ElementId Y equals the requested value;
- every created ID exists and has the expected type/level/location;
- every deleted ID is absent;
- the active view override exists for exactly the intended elements.
After execution, query Revit again. Do not call an echoed input, an execution summary, or the mutating function's own assertion independent readback.
Use these outcome terms consistently:
| Status | Meaning |
|---|---|
verified | Every required claim has independent evidence and passed |
partially_verified | Execution completed, but one or more required claims lack evidence |
verification_failed | At least one required claim was read back and failed |
not_verified | Execution completed without requested readback |
uncertain | Dispatch or timeout makes it unsafe to assert whether execution happened |
Counts are not enough. Compare the actual ElementId sets and claim coverage. If applied count, verified count, and unique IDs disagree, preserve partial, failed, or unknown status.
Show full SKILL.md (632 more words)Show less
7. Report Evidence and Recovery
Return a compact engineering summary containing:
- active document/view binding without exposing sensitive paths;
- target source and exact or redacted target set;
- operation and risk class;
- actual created/modified/deleted counts and IDs when safe to disclose;
- Transaction name and final state;
- verification status, covered claims, failed claims, and missing evidence;
- Undo or rollback instructions;
- any timeout, document change, or unresolved ambiguity.
Use references/evidence-contract.md as the reusable output checklist.
Worked Example
Request: Change the Comments value of the currently selected wall to 待協調.
- Read the active document, view, and selection.
- Stop if selection is empty, has multiple elements for this singular request, or the selected element is not a wall.
- Inspect the selected wall's identity, current Comments value, and parameter writability.
- Preview one target ID, old value, requested value, named Transaction, and Undo availability.
- Execute the parameter write in the supported Revit API context and commit the Transaction.
- Re-query Comments for the same ElementId.
- Report
verifiedonly when the readback equals待協調; otherwise report the observed value andverification_failedoruncertain.
If the call times out after dispatch, do not send the same write again. Re-read the wall first, because the original operation may already have reached the Revit UI thread.
Common Pitfalls
Pitfall: First match becomes the target
- Severity: 🔴 model-impact risk
- Wrong: Mutate the first element returned by a name or category search.
- Right: Present candidates or add deterministic constraints until the exact set is proven.
Pitfall: Success response becomes verification
- Severity: 🔴 false-completion risk
- Wrong: Report the requested value from the execution response as actual model state.
- Right: Perform an independent Revit query and map each required claim to an observation.
Pitfall: Automatic retry after timeout
- Severity: 🔴 duplicate-change risk
- Wrong: Reissue a mutation when completion is unknown.
- Right: Mark the state uncertain, read back the model, then decide whether correction is needed.
Pitfall: Matching counts hide mismatched IDs
- Severity: 🟡 evidence risk
- Wrong: Treat five applied and five verified elements as the same set without comparing IDs.
- Right: Compare unique ID sets and per-claim coverage.
Pitfall: One-target auto-follow becomes a batch prohibition
- Severity: 🟡 workflow risk
- Wrong: Reject every multi-element task.
- Right: Restrict automatic inference; allow explicit, reviewed, bounded multi-target sets.
Evidence Boundaries
| Evidence layer | What it proves | What it does not prove |
|---|---|---|
| Source and policy review | Intended behavior and documented constraints | Compiled or deployed behavior |
| Automated tests | Covered policy branches and data contracts | Live Revit API behavior in the current model |
| Build and deployment checks | Intended artifacts and configuration are present | Correct task result |
| Live smoke | One bounded scenario worked in a named environment | General correctness for other models or versions |
| Independent model readback | Required claims for the current target set | Unchecked adjacent effects |
| Public release/tag | Reproducible version identity and distributable artifact | User-specific installation or live-model state |
Data Currency
- Autodesk Revit API version checked: Revit 2024 documentation.
- Autodesk pages last modified: 2023-10-26; checked 2026-08-28.
- Implementation example: BIM Personal Agent v0.8.1; checked 2026-08-28.
- Volatility: HIGH across Revit versions and agent implementations. Re-check the exact Revit version documentation and the currently loaded add-in/runtime before execution.
To Verify
- Re-check transaction, failure-processing, and ExternalEvent behavior against the exact documentation for any Revit version other than 2024; this entry only checked the 2024 guide.
- Define separate recovery and evidence rules before applying this workflow to linked documents, cross-document plans, or operations that cannot be undone.
Related Skills
- uncertainty-and-source-control — controls certainty labels and source trust; use it when describing unresolved or version-dependent claims.
[^autodesk-revit-2024-transactions]: Autodesk, Revit 2024 API Developers Guide — Transactions. [^autodesk-revit-2024-transaction-classes]: Autodesk, Revit 2024 API Developers Guide — Transaction Classes. [^autodesk-revit-2024-external-events]: Autodesk, Revit 2024 API Developers Guide — External Events. [^bim-personal-agent-v081]: BIM Personal Agent v0.8.1 release, a public personal research project developed by the contributor and used here only as an implementation example; its project-specific policies are not Autodesk requirements.
© h30190, CC-BY-SA-4.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
Files
SKILL.md and 1 other file (references) in raw/設計軟體與工具/BIM工具/Revit/AI目標判定與安全執行/revit-ai-targeting-and-safe-execution of h30190/HJPLUS_Taiwan_Architect_KB.
- SKILL.md
- references/evidence-contract.md
Open the folder on GitHubat commit 35ed06e
Compare with similar skills
Revit AI Targeting And Safe Execution next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| Revit AI Targeting And Safe Execution this skillh30190/HJPLUS_Taiwan_Architect_KB | 158 | — | ~3.6k | Automated safety check: Pass | CC-BY-SA-4.0 | |
| MCP Server Builderanthropics/skills | 180k | 63 repos | ~2.3k | Automated safety check: Pass | Apache-2.0 | |
| MCP Server BuildershareAI-lab/learn-claude-code | 78k | 4 repos | ~1.2k | Automated safety check: Pass | MIT | |
| MCP Integration for Pluginsanthropics/claude-plugins-official | 38k | 11 repos | ~3.1k | Automated safety check: Pass | Apache-2.0 | |
| Microsoft Skill CreatorMicrosoftDocs/mcp | 1.9k | 3 repos | ~2.1k | Automated safety check: Pass | CC-BY-4.0 | |
| Crush Configurationcharmbracelet/crush | 29k | — | ~3.7k | Automated safety check: Pass | Custom licence |
Similar skills
- Official
MCP Server Builder
shareAI-lab/learn-claude-code
Walks through building MCP servers in Python or TypeScript that expose tools, resources and prompts to Claude, with templates, registration and testing.
- Official
MCP Integration for Plugins
anthropics/claude-plugins-official
Explains how to bundle Model Context Protocol servers in a Claude Code plugin, covering config files, stdio, SSE, HTTP and WebSocket server types, and authentication.
- Official
Microsoft Skill Creator
MicrosoftDocs/mcp
Create agent skills for Microsoft technologies using official documentation.
Crush Configuration
charmbracelet/crush
Explains how to configure the Crush coding agent with crushrc or crush.json, covering providers, models, LSPs, MCP servers, hooks, permissions and config precedence.
Context Mode Output Sandbox
mksglu/context-mode
Routes large command, file, API and browser output through context-mode tools so only the needed result enters the agent's context, instead of dumping it via Bash.
More from h30190/HJPLUS_Taiwan_Architect_KB
All 86 skills in this repoGreen Material Search Toolkit
h30190/HJPLUS_Taiwan_Architect_KB
This skill should be used when an architect or consultant needs to search Taiwan's TABC (財團法人臺灣建築中心) green building material certification database, assemble a set of qualified materials for a…
Skill Name Hyphenated
h30190/HJPLUS_Taiwan_Architect_KB
This skill should be used when [specific trigger scenarios].
Taichung Land Parcel Query
h30190/HJPLUS_Taiwan_Architect_KB
This skill should be used when an architect receives just an address or lot number (地號) for a site in Taichung City, Taiwan, and needs the land's basic data — zoning, registered land area, land…
Permit Report Data Mapping
h30190/HJPLUS_Taiwan_Architect_KB
This skill should be used when mapping a legacy CPAMI permit report identifier to its actual BMS data groups, computed display fields, or unresolved template dependencies.
Smoke Exhaust Review
h30190/HJPLUS_Taiwan_Architect_KB
This skill should be used when evaluating smoke exhaust compliance for buildings in Taiwan, including windowless floor determination, smoke compartment partitioning, smoke exhaust window effective…
Accessible Door Clear Width
h30190/HJPLUS_Taiwan_Architect_KB
This skill should be used when checking accessible-route door/opening clear-width compliance for buildings in Taiwan, and avoiding the gap between the code's nominal frame-to-frame distance and the…
Related
Categories
Questions about Revit AI Targeting And Safe Execution
What does Revit AI Targeting And Safe Execution do?
This skill should be used when an AI agent, MCP tool, Revit add-in, or automation must inspect or change a live Revit model and needs to prove its target set, choose bounded read-only context…. Revit AI Targeting And Safe Execution is an agent skill from h30190/HJPLUS_Taiwan_Architect_KB. This skill should be used when an AI agent, MCP tool, Revit add-in, or automation must inspect or change a live Revit model and needs to prove its target set, choose bounded read-only context, preview risk, execute in a supported Revit API context, read the result back independently, and report evidence without overstating success.
When should I use Revit AI Targeting And Safe Execution?
Revit AI Targeting And Safe Execution fits situations like: tasks that involve MCP servers.
How do I install Revit AI Targeting And Safe Execution in Claude Code?
Run `npx skills add h30190/HJPLUS_Taiwan_Architect_KB --skill revit-ai-targeting-and-safe-execution -a claude-code`. Or copy the skill folder (raw/設計軟體與工具/BIM工具/Revit/AI目標判定與安全執行/revit-ai-targeting-and-safe-execution in h30190/HJPLUS_Taiwan_Architect_KB) into .claude/skills/revit-ai-targeting-and-safe-execution in your project. Claude Code loads it when a task matches its description.
How do I install Revit AI Targeting And Safe Execution in Codex?
Run `npx skills add h30190/HJPLUS_Taiwan_Architect_KB --skill revit-ai-targeting-and-safe-execution -a codex`. Or copy the skill folder (raw/設計軟體與工具/BIM工具/Revit/AI目標判定與安全執行/revit-ai-targeting-and-safe-execution in h30190/HJPLUS_Taiwan_Architect_KB) into .agents/skills/revit-ai-targeting-and-safe-execution in your project. Codex loads it when a task matches its description.
Can I use Revit AI Targeting And Safe Execution in Cursor, Gemini CLI or GitHub Copilot?
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add h30190/HJPLUS_Taiwan_Architect_KB --skill revit-ai-targeting-and-safe-execution -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/revit-ai-targeting-and-safe-execution, .gemini/skills/revit-ai-targeting-and-safe-execution, .github/skills/revit-ai-targeting-and-safe-execution and .opencode/skills/revit-ai-targeting-and-safe-execution in your project.
What does Revit AI Targeting And Safe Execution need to run?
SKILL.md names no scripts, command-line tools or credentials: Revit AI Targeting And Safe Execution is instructions for the agent only. Compatibility (from SKILL.md): claude-code,opencode,agent-skills.
Does Revit AI Targeting And Safe Execution access the network?
SKILL.md names 2 domains. As links in the text: help.autodesk.com and github.com. This is read from the text; nothing was executed.
Is Revit AI Targeting And Safe Execution safe to install?
Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.
What licence does Revit AI Targeting And Safe Execution use?
Revit AI Targeting And Safe Execution is published under the CC-BY-SA-4.0 licence (declared in SKILL.md). It allows redistribution, so the full SKILL.md is shown on this page.
How many tokens does Revit AI Targeting And Safe Execution use?
About 3.6k tokens (SKILL.md is roughly 14k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 904 tokens, read only when the agent opens those files.
What are the alternatives to Revit AI Targeting And Safe Execution?
Skills that share tags, products or a category with Revit AI Targeting And Safe Execution: MCP Server Builder (anthropics/skills, 180k stars), MCP Server Builder (shareAI-lab/learn-claude-code, 78k stars), MCP Integration for Plugins (anthropics/claude-plugins-official, 38k stars) and Microsoft Skill Creator (MicrosoftDocs/mcp, 1.9k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
Who maintains Revit AI Targeting And Safe Execution?
h30190 (a GitHub user) maintains it in h30190/HJPLUS_Taiwan_Architect_KB, which has 158 GitHub stars. The repository holds 86 skills in this directory. The repository was last updated on October 8, 2026.
Source: h30190/HJPLUS_Taiwan_Architect_KB on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.