> ## Documentation Index
> Fetch the complete documentation index at: https://docs.sixtyfour.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Team skills

> Give your team's research agents playbooks: they read one when it fits, or follow it for a whole run.

A team skill is a short Markdown playbook, such as how your team confirms where a person worked. Each deep research run lists the team's skills, and an agent reads one when its description fits the task. A run can also attach skills, so every agent follows them from its first step.

A skill guides Sixtyfour's research agents and never changes their tools or budget. To teach your own coding assistant to call Sixtyfour, see [Skills for AI Agents](https://docs.sixtyfour.ai/developer-tools/skills.md).

> [!NOTE]
>
> Team skills are enabled per organization. Contact [support](mailto:support@sixtyfour.ai) to turn them on.

## Write a skill

```markdown theme={null}
---
name: employment-history
description: Use when confirming where a person worked and when.
---

1. Start with company filings and press releases that name the person and a title.
2. Treat a profile's dates as a claim to confirm, not as evidence.
3. Record each employer with the source that dates it.
```

Agents decide from the description alone, so say when to use the skill, not only what it covers.

## Set it up

In the app, open **Settings → Agents → Skills** and choose the **Team** tab. Every member can read the team's skills. Team owners add, change and delete them, and on the default team, organization admins and owners do.

Through the API, use an API key for the team:

| Request | What it does |
| - | - |
| `GET /skills` | List the team's skills, and whether you can change them. |
| `GET /skills/{name}` | Read one skill. |
| `PUT /skills/{name}` | Create a skill from `markdown`, or replace it by also sending the `version` a read returned. |
| `DELETE /skills/{name}?version=...` | Delete a skill. |

```bash theme={null}
curl -X PUT "https://api.sixtyfour.ai/skills/employment-history" \
  -H "x-api-key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"markdown": "---\nname: employment-history\ndescription: Use when confirming where a person worked and when.\n---\n\nStart with company filings and press releases.\n"}'
```

The response is the saved skill. Send its `version` to replace or delete it.

```json theme={null}
{
  "name": "employment-history",
  "description": "Use when confirming where a person worked and when.",
  "markdown": "---\nname: employment-history\ndescription: Use when confirming where a person worked and when.\n---\n\nStart with company filings and press releases.\n",
  "version": "2026-11-03T18:04:11.512304+00:00"
}
```

## Use it in a run

Runs on `xhigh`, on `scout` with `graph: true`, and on `high` when enabled for your organization list the team's skills. To make every agent of a run follow a skill, attach it with `skills` on [`POST /people-intelligence-async`](https://docs.sixtyfour.ai/atlas/integrations#request-parameters):

```bash theme={null}
curl -X POST "https://api.sixtyfour.ai/people-intelligence-async" \
  -H "x-api-key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"lead_info": {"name": "Jane Doe", "company": "Acme"}, "tier": "xhigh", "skills": ["team/employment-history"]}'
```

Every agent of the run holds an attached skill's whole text from its first turn, at the version the run started with. The [MCP server](https://docs.sixtyfour.ai/developer-tools/mcp-setup.md)'s `sixtyfour_atlas_start` takes the same `skills`. The Atlas API's start and continue endpoints don't.

## Limits

| | Limit |
| - | - |
| Name | Lowercase letters and digits joined by hyphens, up to 64 characters |
| Description | 500 characters |
| Instructions | 20,000 characters |
| Whole file | 24,000 characters |
| The team's listing, one line per skill | 2,000 tokens, about 8,000 characters |
| Skills attached to one run | 25,000 tokens, about 100,000 characters |

Requests to `/skills` share one [rate limit](https://docs.sixtyfour.ai/api-reference/rate-limits.md#skills).

## Good to know

* A run keeps the list it started with. Reading a listed skill that isn't attached returns its latest saved text, and one deleted since is refused.
* The frontmatter holds `name` and `description`. A `license` or `metadata` field, as Claude Code and Codex skills carry, is dropped. Any other field is refused.
* A skill is one file. Link a full `https://` address, not a file beside it.
* Saving with an old `version` saves nothing. Read the skill again and reapply your change.
* A run that attaches an unknown skill or too much text is refused before it starts or is charged.
* An API key acts with its creator's current role in the team. See [Roles and permissions](https://docs.sixtyfour.ai/guides/roles-and-permissions.md).

## Errors

For error responses (400, 403, etc.), see [Handling Errors](https://docs.sixtyfour.ai/api-reference/errors.md#skills).


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.