# Postey Teams

> >

- **Type:** Skill
- **Install:** `agentstack add skill-posteyai-skills-postey-teams`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [posteyai](https://agentstack.voostack.com/s/posteyai)
- **Installs:** 0
- **Category:** [Communication](https://agentstack.voostack.com/c/communication)
- **Latest version:** 0.1.0
- **License:** MIT
- **Upstream author:** [posteyai](https://github.com/posteyai)
- **Source:** https://github.com/posteyai/skills/tree/main/skills/postey-teams
- **Website:** https://postey.ai

## Install

```sh
agentstack add skill-posteyai-skills-postey-teams
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# Postey Teams

## Approval is a convention, not a state

**Postey has no `APPROVED` status.** `PostStatusEnum` is `DRAFT | SCHEDULED | PUBLISHING |
PUBLISHED`. Nothing in the system records that a human said yes.

That has three consequences, and getting them wrong is how content goes out unreviewed:

| What is true | What it means for you | The failure it prevents |
|---|---|---|
| A draft carries no approval flag | Approval lives in an internal comment, a message, or a conversation. Read it; never infer it. | Treating a status as consent |
| Silence is not approval | If asked whether something is cleared, say who said what and when — never "no objections were raised" | Reporting absence of dissent as sign-off |
| Nothing stops a `DRAFT` publishing | The status does not protect it. Only the rule that publishing needs explicit instruction does, and that rule lives in the agent, not the server. | Assuming the server will catch you |

When the user says "it's approved", that is the approval. Take it from them, and say whose approval
you are acting on.

## Internal comments are the review record

`postey://posts/{post_id}/comments/{platform}` holds team notes — visible inside Postey, never
published. They work on a post in any status, not just drafts.

Read them **before** drafting a revision. The most common failure in a review loop is rewriting a
post while ignoring the note that says what was wrong with it.

Internal comments are not audience comments. Confusing the two means treating a colleague's note
as public feedback — or worse, drafting a public reply to an internal note.

| | Internal comments | Audience comments |
|---|---|---|
| **Who wrote it** | A teammate or client, inside Postey | A real person on the live post |
| **Reach** | Never published | Public |
| **Read via** | `postey://posts/{post_id}/comments/{platform}` | `get_platform_comments` |
| **Owned by** | This skill | `postey-engagement` |
| **A reply goes** | Back into Postey | Out to the public |

## Share links

`resolve_share_link` resolves a Postey share URL or `share_id` to the underlying post. Use it
when someone pastes a link — the reviewer often has only that.

A share link is how a client who has no Postey account sees a draft. Two things follow: the
recipient may not be in the team list, and anyone holding the link can read the draft. Do not put
anything in a draft you would not send to whoever might be forwarded that link.

## Teams

`postey://teams` and `postey://teams/{team_id}/members` say who exists and who can act. Read them
before attributing a comment or asking "who should approve this?" — guessing a reviewer's name in
front of a client is a small error with an outsized cost.

An account can belong to several teams, and identities can be shared across accounts
(`shares_identity_with` in `postey://accounts`). Two account entries pointing at the same underlying
identity will publish to the same place. Check before assuming two accounts mean two audiences.

## The routine that works

1. Read the accounts and confirm which one this content is for.
2. Draft. Everything stays a `DRAFT`.
3. Hand over the share link and say explicitly who needs to approve.
4. When they respond, read the internal comments and the user's own words.
5. Revise against what was actually said, and show the diff.
6. Publish or schedule **only** on explicit instruction — and scheduling counts as publishing.

Step 6 does not become optional because step 3 happened. A reviewer approving a draft is not the
user telling you to publish it.

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [posteyai](https://github.com/posteyai)
- **Source:** [posteyai/skills](https://github.com/posteyai/skills)
- **License:** MIT
- **Homepage:** https://postey.ai

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-posteyai-skills-postey-teams
- Seller: https://agentstack.voostack.com/s/posteyai
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.
