For agents and technical readers

Discover a skill and prepare a tested revision

Use the Mind's current skill discovery to select a procedure the connected client can perform. Treat authoring and executing a skill as separate work.

Contents

Discover and load

Call orient, then list_skills with {}. The optional capability filter accepts any or repo-agent. Read the chosen procedure with get_skill:

{"name":"<discovered skill name>"}

Inspect its surface and runtime requirements. If the client cannot run a required script, repository step or subagent, report that limitation and hand the work to a compatible client. Do not treat reading the skill as successful execution.

Follow the procedure within authority

Read its relevant supporting references and source records. Use the current client-facing schemas for each called tool. Keep the user's output and external-action authority separate from what the procedure recommends. Verify the resulting saved record or output and cite the source actually read.

Author from observed work

For a repeated job, load get_skill with {"name":"skill-engineer"}. Check for an existing skill, gather real corrections and define outcome and trigger cases before drafting. A rule that must always hold belongs in a deterministic gate; fixed repeatable steps may belong in a script; an independently scheduled program belongs in an app bundle.

Run the procedure's baseline and candidate checks in the intended client and model within the agreed spend. Record observed outcomes; an unrun case is unverified. Propose the relevant instruction files through the authorised governed route and inspect the result. Instructions being saved does not prove later execution.

Retire or revise deliberately

When a tool or task changes, use the same authoring procedure to revise the skill and rerun affected cases. Keep one clear procedure rather than near-duplicate instructions. Report unsupported or unavailable authoring steps without claiming registration or publication.

Related

Use a skill or turn a repeated job into one · Build, propose and verify a governed model agent

Practical guides for people and agents

Start with one useful job

Join the waitlist

Plain text for AI agents. Same content as the human page.

## For agents

- What Mensara is: it gives a team its own Mind, one shared memory for the team and their AI agents, built by Xanadu Labs LLC.
- Status: signup is not open. A person joins the waitlist at https://app.mensara.ai/#waitlist. Agents cannot create an account or sign up.
- How you can help your person: explain what Mensara does, check whether it fits a job their team repeats, and share the waitlist link.
- Agent references: https://app.mensara.ai/llms.txt lists every page as Markdown and the tool input references for connected agents.

---

# Discover a skill and prepare a tested revision

Operating procedures · For agents and technical readers · [How it works](https://app.mensara.ai/how-it-works)

Use the Mind's current skill discovery to select a procedure the connected client can perform. Treat authoring and executing a skill as separate work.

[Page as Markdown](https://app.mensara.ai/how-it-works/skill-procedure.md)

[Human guide](https://app.mensara.ai/how-it-works/use-and-create-skills)

## Use the released interface with current authority

Availability: Core0.20.11 / Portal0.3.47 source checked; discover the actual installed client and enabled features.

Who can do this: The connected principal with the permissions named in this procedure

What changes: The procedure explains its actual reads, writes or external effects before action.

Limits: An input accepted by a source schema is not a successful installed-runtime task.; Keep human decisions, client exclusions and pending results explicit.

Verification: Source checked and independently reviewed; scoped runtime receipts are recorded separately.

## Discover and load

Call `orient`, then `list_skills` with `{}`. The optional `capability` filter accepts `any` or `repo-agent`. Read the chosen procedure with `get_skill`:

```
{"name":"<discovered skill name>"}
```

Inspect its surface and runtime requirements. If the client cannot run a required script, repository step or subagent, report that limitation and hand the work to a compatible client. Do not treat reading the skill as successful execution.

## Follow the procedure within authority

Read its relevant supporting references and source records. Use the current client-facing schemas for each called tool. Keep the user's output and external-action authority separate from what the procedure recommends. Verify the resulting saved record or output and cite the source actually read.

## Author from observed work

For a repeated job, load `get_skill` with `{"name":"skill-engineer"}`. Check for an existing skill, gather real corrections and define outcome and trigger cases before drafting. A rule that must always hold belongs in a deterministic gate; fixed repeatable steps may belong in a script; an independently scheduled program belongs in an app bundle.

Run the procedure's baseline and candidate checks in the intended client and model within the agreed spend. Record observed outcomes; an unrun case is unverified. Propose the relevant instruction files through the authorised governed route and inspect the result. Instructions being saved does not prove later execution.

## Retire or revise deliberately

When a tool or task changes, use the same authoring procedure to revise the skill and rerun affected cases. Keep one clear procedure rather than near-duplicate instructions. Report unsupported or unavailable authoring steps without claiming registration or publication.

## Related

- [Use a skill or turn a repeated job into one](https://app.mensara.ai/how-it-works/use-and-create-skills)
- [Build, propose and verify a governed model agent](https://app.mensara.ai/how-it-works/build-agent-procedure)

[Join the waitlist](https://app.mensara.ai/#waitlist)