agentleFS
kjendrzycakjendrzyca/public-skills/skills/tally/SKILL.md

tally

Skill1 starsChanged 3 months agoMITView on GitHub

Create and edit Tally.so forms and read responses.

2 other projects use a version of this.

Adopted by

Most recently taken up by

The file

Show its rules and text

We split this skill into the rules it gives an agent

14 rules. About 0 are common, 5 are distinctive, and 9 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 9

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Use the official Tally MCP server through MCPorter.

    line 9
  2. Distinctive:

    Read compact tool signatures:

    line 18
  3. Distinctive:

    Call the tool with arguments matching its current schema:

    line 33
  4. Distinctive:

    If authentication is missing, follow connection.md.

    line 43
  5. Distinctive:

    Do not dump all schemas or load all reference pages.

    line 44
  6. Very specific:

    Before create_new_form, load_form, or any operation on the working form, read mcporter-daemon.md and acquire its session lock.

    line 49
  7. Very specific:

    Start with create_new_form or load_form.

    line 53
  8. Very specific:

    Create the form the user requested.

    line 55
  9. Very specific:

    Always pass save_form.status explicitly: its current default is PUBLISHED.

    line 58
  10. Very specific:

    After the first save, always pass the saved formId on subsequent saves.

    line 59
  11. Distinctive:

    Before editing, load the target form and retain its existing IDs.

    line 60
  12. Very specific:

    If the intended title ends with (...), restore the full title using update_text after creation.

    line 65
  13. Very specific:

    Before saving, call list_blocks and compare the working content with the request.

    line 67
  14. Very specific:

    Read them on demand through MCPorter.

    line 84

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Tally

Use the official Tally MCP server through MCPorter. Resolve mcporter.json relative to this SKILL.md, and set TALLY_MCP_CONFIG to that resolved absolute path in each shell invocation (or export it in a persistent shell). Do not infer it from the working directory. Examples below assume this variable is set. The bundled config uses OAuth, keeps the connection alive, and disables imports from other MCP clients. See connection.md for setup.

Discover only what you need

  1. Read compact tool signatures:

    mcporter --config "$TALLY_MCP_CONFIG" list tally --brief --no-oauth
  2. Pick the exact tool name from that output, then inspect only its schema. For large tools (create_blocks, configure_blocks, update_settings), use filtered discovery to read only the needed block types, operations, or settings. For small tools:

    mcporter --config "$TALLY_MCP_CONFIG" list tally.TOOL_NAME --schema --no-oauth
  3. Call the tool with arguments matching its current schema:

    mcporter --config "$TALLY_MCP_CONFIG" call tally.TOOL_NAME --args '{"ARGUMENT":"VALUE"}' --output json --no-oauth

    TOOL_NAME, ARGUMENT, and VALUE are placeholders. Do not guess names or reuse arguments from a different tool. For large payloads, use a local JSON file and shell-quote its contents: --args "$(cat /absolute/path/arguments.json)".

If authentication is missing, follow connection.md. Do not dump all schemas or load all reference pages. MCP tools may themselves return usage instructions; follow those for the chosen operation.

Forms

  • Before create_new_form, load_form, or any operation on the working form, read mcporter-daemon.md and acquire its session lock. The keep-alive connection is shared by agents and holds one working form. Hold the lock through editing, saving, and verification; never interleave forms.
  • Start with create_new_form or load_form. Edit through the returned ledger and targeted MCP tools. These changes remain in memory until save_form.
  • Create the form the user requested. Use DRAFT unless publishing was requested; preserve an existing form's status unless the task changes it. If a tool cannot express the requested status, explain that before writing.
  • Always pass save_form.status explicitly: its current default is PUBLISHED. After the first save, always pass the saved formId on subsequent saves.
  • Before editing, load the target form and retain its existing IDs. Use UUIDs from the returned ledger; MCP creation tools generate new IDs.
  • create_blocks currently makes questions required. Use configure_blocks to make requested fields optional, then verify the ledger.
  • create_blocks may strip the final parenthesized phrase from a question title. If the intended title ends with (...), restore the full title using update_text after creation. Verify it after saving and reloading the form.
  • Before saving, call list_blocks and compare the working content with the request. After saving, reload the saved form and check its content; verify status through saved metadata or list_forms. Do not reload during unsaved editing, because loading replaces the working form. Report its ID/link and whether it is draft or published. If a save times out, inspect stored state before retrying; do not blindly create a duplicate.

Responses

Inspect fetch_submissions for response retrieval or fetch_insights for aggregate analytics. Filter to the requested form, period, and completion state; follow pagination when the task needs the complete set. Report the scope and number of responses covered. Keep answers out of the skill and committed fixtures.

Further details

Tool descriptions and live schemas define the supported operations and arguments. Read them on demand through MCPorter. If they leave a product question unanswered, inspect and use Tally's search_documentation tool. Report missing capabilities; do not invent tools or silently switch to another interface.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.