AGENT SKILL

ogment-agentic-cli

暂无简介

License 许可
MIT
最近更新
2026-06-07
Stars
★ 0
Forks
1

03 SKILL.md 正文

Ogment CLI Skill

Securely invoke MCP tools via the Ogment CLI. Access your connected SaaS tools (Linear, Notion, Gmail, PostHog, etc.) through Ogment's governance layer.


Quick Start (First-Time Onboarding)

Follow this flow when first using Ogment with a user:

Step 1: Check Auth

ogment auth status
  • If loggedIn: true → skip to Step 3
  • If loggedIn: false → continue to Step 2

Step 2: Login (if needed)

ogment auth login

Extract the code from the response and send it to your human.

⚠️ Make links clickable! Use markdown or full URLs so humans can tap/click directly.

🔐 Approve this code to connect Ogment: XXXX-XXXX 👉 dashboard.ogment.ai/cli/approve

Wait for approval, then verify with ogment auth status.

Step 3: Discover What's Available

ogment catalog

Then for each server:

ogment catalog <serverId> | jq '[.data.tools[].name]'

Step 4: Summarize to Your Human

Tell them what you found:

✅ Connected to Ogment! Here's what I can access:

  • Linear: 28 tools (issues, projects, teams, docs)
  • Gmail: 11 tools (messages, threads, drafts)
  • Notion: 5 tools (search, fetch, comments)
  • Slack: 7 tools (conversations, users)

What would you like me to help with?


Prerequisites

Requirement Install Required
ogment CLI npm install -g @ogment-ai/cli ✅ Yes
jq brew install jq / apt install jq Optional (for filtering)

First-Time Setup (Login Flow)

⚠️ IMPORTANT FOR AGENTS: Don't tell the human to run ogment auth login — run it yourself and send them the code!

Step 1: Check if already authenticated

ogment auth status

If loggedIn: true, skip to Core Workflow.

Step 2: If not logged in, start device flow

ogment auth login

This returns JSON with a device code. Extract and send to the human:

Example output:

{
  "data": {
    "event": "auth_login.pending",
    "verification": {
      "userCode": "ABCD-1234",
      "verificationUri": "https://dashboard.ogment.ai/cli/approve"
    }
  }
}

Step 3: Send the code to your human

Tell them:

Approve this code: ABCD-1234 👉 https://dashboard.ogment.ai/cli/approve

Step 4: Wait for approval

The ogment auth login command will complete automatically once approved. Then verify:

ogment auth status

Authentication & Credentials

  • Credentials location: ~/.config/ogment/credentials.json
  • Token management: Ogment handles OAuth for all connected services
  • Scope: Access depends on services connected in your Ogment dashboard
  • Per-agent permissions: Each agent only sees tools you've explicitly granted

No credentials are stored in this skill — all auth is managed by the Ogment CLI.

When to Use

  • User asks to interact with their connected services (issues, docs, emails, analytics)
  • You need to call MCP tools that require auth/credentials
  • Discovering what integrations the user has available

Core Workflow

status → catalog → catalog <server> → catalog <server> <tool> → invoke

1. Check connectivity (if issues suspected)

ogment status

Returns auth state, connectivity, and available servers. Check summary.status for quick health.

2. Discover servers

ogment catalog

Returns list of servers with serverId and toolCount. Use serverId in subsequent calls.

3. List tools on a server

ogment catalog <serverId>

Returns all tools with name and description. Scan descriptions to find the right tool.

4. Inspect tool schema

ogment catalog <serverId> <toolName>

Returns inputSchema with properties, types, required fields, and descriptions.

5. Invoke a tool

ogment invoke <serverId>/<toolName> --input '<json>'

Input methods:

  • Inline JSON: --input '{"query": "test"}'
  • File: --input @path/to/input.json
  • Stdin: echo '{}' | ogment invoke ... --input -

6. Debug errors

ogment invoke <serverId>/<toolName> --input '{}' --debug

The --debug flag surfaces raw MCP error messages with field-level validation details.

Security Considerations

File Input Safety

The --input @path flag reads local files. Avoid these paths:

  • ~/.ssh/* — SSH keys
  • ~/.aws/* — AWS credentials
  • ~/.config/ — App configs and tokens
  • ~/.bash_history, ~/.zsh_history — Shell history
  • Browser profile directories

Best practice: Only use --input @path with files you've explicitly created for this purpose.

Network Security

  • All API calls route through dashboard.ogment.ai
  • No direct connections to SaaS APIs
  • TLS encrypted in transit

Permission Model

  • Tools are scoped per-agent in your Ogment dashboard
  • Agents only see tools you've granted access to
  • Write operations may be restricted based on agent permissions

Output Format

All commands return structured JSON:

{
  "ok": true,
  "data": { ... },
  "error": null,
  "meta": { "command": "..." },
  "next_actions": [
    { "command": "...", "title": "...", "reason": "..." }
  ]
}
  • Check ok first — boolean success indicator
  • next_actions — suggested follow-up commands
  • error.category — validation, not_found, remote, auth, internal
  • error.retryable — whether retry might help

Common Patterns

Find a tool by intent

ogment catalog <serverId> | jq '.data.tools[] | select(.name + .description | test("email"; "i"))'

List issues assigned to user

ogment invoke openclaw/Linear_list_issues --input '{"assignee": "me"}'

Search Notion

ogment invoke openclaw/Notion_notion-search --input '{"query": "quarterly review", "query_type": "internal"}'

Get Gmail messages

ogment invoke openclaw/gmail_listMessages --input '{"q": "is:unread", "maxResults": 10}'

Error Recovery

Error Code Meaning Action
TOOL_NOT_FOUND Bad server/tool name Run ogment catalog to rediscover
VALIDATION_INVALID_INPUT Malformed JSON Check JSON syntax
TRANSPORT_REQUEST_FAILED Server rejected call Add --debug, check schema
AUTH_INVALID_CREDENTIALS Bad/expired API key Run ogment auth login
HTTP_401 Service connection expired Tell human to reconnect (see below)
HTTP_502 Server down Retry after delay

Handling Expired Connections

When you get HTTP_401 with a message like:

"Your connection to [Service] has expired. Please reconnect..."

Tell your human (with clickable link):

⚠️ Your [Service] connection has expired. Please reconnect it here: dashboard.ogment.ai (Go to Integrations → [Service] → Reconnect)

Let me know when done and I'll retry!

Handling Missing Permissions

If a tool you expect isn't available (e.g., gmail_createDraft not in catalog):

  • This is normal — agents have scoped permissions
  • Write tools may be disabled by default

Tell your human (with clickable link):

I don't have write access to [Service]. To enable it, go to: dashboard.ogment.ai (Agents → [Agent Name] → Permissions)

Let me know when updated and I'll check again!

Exit Codes

Code Meaning
0 Success
2 Validation/parse error
5 Not found
7 Remote/transport error
9 Internal error

Flags Reference

Flag Effect
--debug Include raw error diagnostics
--human Human-readable output
--yes Auto-confirm prompts
--api-key <key> Override API key

Avoid: --quiet (suppresses all output including data)

Pre-flight Checklist

Before invoking a tool:

  1. ✅ Confirmed server exists (catalog)
  2. ✅ Confirmed tool exists (catalog <server>)
  3. ✅ Checked required fields in schema
  4. ✅ Matched types exactly (number vs string)
  5. ✅ Used exact casing for IDs
  6. ✅ If using --input @path, verified file doesn't contain sensitive data

04 同仓库更多 Skills

javimosch
javimosch/open-claw-skills

crowd-prompting

A marketplace where AI agents improve prompts, system instructions, tool descriptions, and other text-based content with domain expertise from real-world operations — and earn tokens for valuable contributions.

★ 0
javimosch
javimosch/open-claw-skills

ceaser

Interact with the Ceaser privacy protocol on Base L2 using the ceaser-mcp MCP tools. This skill uses the ceaser-mcp npm package for ALL operations -- shield, unshield, note management, and protocol queries. All ceaser tool calls use CLI subcommands (npx -y ceaser-mcp <subcommand>). Ceaser is a privacy-preserving ETH wrapper using Noir/UltraHonk zero-knowledge proofs on Base L2 (chain ID 8453).

★ 0
javimosch
javimosch/open-claw-skills

ceaser-send

Fully automated private ETH transfer via Ceaser Protocol on Base L2 using the ceaser-mcp MCP tools. This skill uses the ceaser-mcp npm package for ALL operations -- all ceaser tool calls use CLI subcommands (npx -y ceaser-mcp <subcommand>). WARNING -- creates observable on-chain links between user wallet and hot wallet (see Privacy Warning). Generates an ephemeral hot wallet with BIP-39 mnemonic recovery, user funds it, agent signs and broadcasts the Shield TX automatically. Exactly one manual step required (funding the hot wallet with ETH). Uses Noir/UltraHonk zero-knowledge proofs.

★ 0
javimosch
javimosch/open-claw-skills

dilemm-ai

Compete in DilemmAI, the prisoner's dilemma AI arena at dilemm.ai. Use when an OpenClaw agent wants to create an account, design and submit strategy prompts for their bot, enter matchmaking, analyze historical matches and leaderboards, and iterate on bot strategy. Requires headless browser automation (no API available yet).

★ 0
javimosch
javimosch/open-claw-skills

moldium

Post and manage content on the Moldium blog platform. Triggered by "post to Moldium", "write a blog post", "publish an article", etc.

★ 0
javimosch
javimosch/open-claw-skills

arxiv-paper-reviews

Interact with arXiv Crawler API to fetch papers, read reviews, submit comments, search papers, and import papers. Use when working with arXiv papers, fetching paper lists by date/category/interest, viewing paper details with comments, submitting paper reviews, searching papers by title, or importing papers from arXiv URLs via API at http://weakaccept.top:8000/.

★ 0