Bankr Agent Job Workflow
- 370 installs
- 80 repo stars
- Updated March 24, 2026
- bankrbot/claude-plugins
bankr agent - job workflow is a Claude Code skill that teaches the submit-poll-complete asynchronous job pattern for Bankr MCP tools so developers who integrate crypto agents can reliably track multi-step API operations.
About
bankr agent - job workflow (version 1.0.0) in BankrBot/claude-plugins documents the core asynchronous pattern for every Bankr API operation through three MCP tools: bankr_agent_submit_prompt, bankr_agent_get_job_status, and bankr_agent_cancel_job. Agents submit natural-language prompts, receive a job ID, poll every 2 seconds, and read terminal states pending, processing, completed, failed, or cancelled along with transactions and statusUpdates. Typical jobs finish in 30 seconds to 2 minutes; the skill recommends suggesting cancellation after 3+ minutes for simple queries. Output guidelines cover price quotes, trades, Polymarket odds, balances, and error recovery when polling fails but the server-side job continues. The skill ships inside the bankr-agent plugin alongside 15 companion skills for trading, Polymarket, token deployment, and safety. Developers reach for bankr agent - job workflow when wiring any Bankr MCP integration or debugging stuck jobs. Triggers include Bankr API requests, job status checks, and submit-poll-complete workflow questions.
- Multi-step job state machines
- Retry and failure handling
- Human-in-the-loop checkpoints
- Bankr task orchestration patterns
Bankr Agent Job Workflow by the numbers
- 370 all-time installs (skills.sh)
- Ranked #2,068 of 16,546 AI & Agent Building skills by installs in the Skillselion catalog
- Data as of Jul 31, 2026 (Skillselion catalog sync)
npx skills add https://github.com/bankrbot/claude-plugins --skill bankr-agent---job-workflowAdd your badge
Show developers this skill is listed on Skillselion. Paste this into your README.
| Installs | 370 |
|---|---|
| repo stars | ★ 80 |
| Last updated | March 24, 2026 |
| Repository | bankrbot/claude-plugins ↗ |
How do Bankr agent API jobs submit and poll?
Model multi-step Bankr jobs as durable workflows with statuses, retries, and handoffs between agent tasks and human approval gates.
Who is it for?
Developers integrating Bankr MCP tools into Claude Code or Cursor agents who need a reliable async job pattern for crypto API calls.
Skip if: Teams configuring Bankr API key security or wallet access controls without executing agent jobs through MCP.
When should I use this skill?
User submits Bankr prompts, polls job status, uses Bankr MCP tools, or asks about the submit-poll-complete workflow pattern.
What you get
Tracked job IDs, terminal status responses, executed transaction arrays, and formatted agent output for prices, trades, or balances.
- Completed job responses
- Transaction arrays
- Formatted agent trade or balance output
By the numbers
- 3 MCP tools for submit, poll, and cancel
- 5 job status states with 2-second poll interval
- Part of bankr-agent plugin with 16 companion skills
Files
Safety & Access Control
Comprehensive safety guidance for using the Bankr API and CLI.
API Key Types
Bankr uses a single key format (bk_...) with capability flags:
| Flag | Controls | Default |
|---|---|---|
agentApiEnabled | /agent/* endpoints | false |
llmGatewayEnabled | LLM Gateway at llm.bankr.bot | false |
readOnly | Restricts agent to read-only tools | false |
Manage flags at bankr.bot/api.
Read-Only API Keys
When readOnly: true, all write tools are filtered from the agent session:
Allowed: Balances, prices, analytics, portfolio, research Blocked: Swaps, transfers, NFT purchases, staking, orders, token launches, leverage, Polymarket bets, claims
The /agent/sign and /agent/submit endpoints return 403 for read-only keys.
IP Whitelisting
Set allowedIps on your API key to restrict usage to specific IPs. Requests from non-whitelisted IPs are rejected with 403.
- Empty array = all IPs allowed (default)
- Non-empty array = only listed IPs can use the key
Dedicated Agent Wallet
When building autonomous agents, create a separate Bankr account rather than using your personal wallet:
- Limited exposure — compromised key only exposes agent wallet funds
- Clear accounting — agent transactions isolated from personal activity
- Independent controls — apply stricter access controls without affecting personal use
- Easy revocation — disable agent account without disrupting primary wallet
Recommended Funding
| Chain | Gas Buffer | Trading Capital |
|---|---|---|
| Base | 0.01-0.05 ETH | As needed |
| Polygon | 5-10 MATIC | As needed |
| Ethereum | 0.05-0.1 ETH | As needed |
| Solana | 0.1-0.5 SOL | As needed |
Rate Limits
| Tier | Daily Limit |
|---|---|
| Standard | 100 messages/day |
| Bankr Club | 1,000 messages/day |
| Custom | Set per API key |
Reset window is 24 hours from first message (rolling), not midnight.
Key Management
- Store keys in environment variables (
BANKR_API_KEY,BANKR_LLM_KEY), never in source code - Add
~/.bankr/and.envto.gitignore - Rotate keys periodically at bankr.bot/api
- Revoke immediately if compromised
- Use separate keys for Agent API vs LLM Gateway if needed
Transaction Safety
- Test first — small amounts on Base/Polygon before scaling
- Verify recipients — double-check addresses before transfers
- Gas buffer — keep enough native tokens for gas
- Immediate execution —
/agent/submitexecutes with no confirmation prompt - Understand calldata — verify trusted source for arbitrary transactions
Pre-Deployment Checklist
- Use a dedicated agent wallet, not your personal account
- Fund with limited amounts appropriate to purpose
- Set read-only if agent only queries data
- Configure IP whitelisting for server-side agents
- Store keys in environment variables
- Test with small amounts first
- Implement error handling for rate limits (429) and access control (403)
Related skills
Setup Matt Pocock SkillsScaffold the per-repo configuration that Matt Pocock’s engineering agent skills rely on so they understand the issue tracker, triage labels, and domain documentation la495k196k
Lark AppsConnect Claude, Cursor or custom agents directly to Lark (Feishu) for messaging, document automation, approval workflows and enterprise data access.402k
CavemanSlash token usage by roughly 75% while keeping every technical detail intact when working with Claude Code, Cursor or similar agents.391k94.7k
Lark Skill MakerQuickly turn any Lark/Feishu OpenAPI call or multi-step workflow into a reusable agent skill with its own SKILL.md.386k16k
Running Claude Code Via Litellm CopilotRun Claude Code at a fraction of the cost by routing requests through LiteLLM to the GitHub Copilot Chat API.270k72
Codex PetGenerate a complete Codex Pet spritesheet and metadata from one reference image without needing an OpenAI key or Codex Pro.246k10How it compares
Pick bankr agent - job workflow for the core async MCP pattern; pick bankr-safety for API key and wallet access control configuration.
FAQ
What MCP tools does bankr agent - job workflow use?
bankr agent - job workflow centers on three MCP tools: bankr_agent_submit_prompt to start jobs, bankr_agent_get_job_status to poll every 2 seconds, and bankr_agent_cancel_job to stop running work.
How long should agents poll Bankr jobs?
bankr agent - job workflow specifies a 2-second poll interval with typical completion in 30 seconds to 2 minutes, and recommends suggesting cancellation after 3+ minutes for simple queries.
What job statuses does Bankr return?
bankr agent - job workflow documents five terminal and in-flight states—pending, processing, completed, failed, and cancelled—each with defined agent actions for polling, reporting, or error handling.