agent-wallet-policy
View spending policy on a Circle agent wallet — per-transaction, daily, weekly, and monthly USDC caps via the `circle` CLI. Use when the user wants to inspect current limits. Setting or resetting limits requires OTP confirmation in an interactive terminal session — the agent hands the user a verbatim command to run themselves; the OTP must never pass through agent storage. Mainnet-only — testnet chains are rejected. Triggers on: spending limit, spending policy, wallet limit, per-tx cap, daily cap, weekly cap, monthly cap, set spending limit, reset spending limit, wallet rules, spending cap, OTP confirmation.
What this skill does
## Overview For an overview of the Circle CLI's **full** capability set, see the `use-circle-cli` master skill. This skill is the narrower spending-policy flow. Circle agent wallets support **spending policies** — per-wallet caps that the CLI enforces on every payment and transfer. There are three operations: | Operation | Command | OTP required? | |---|---|---| | **View** current limits | `circle wallet limit --address <addr> --chain BASE --output json` | No | | **Set** custom limits | `circle wallet limit set ...` | **Yes — human OTP, run in user's own terminal** | | **Reset** to defaults | `circle wallet limit reset ...` | **Yes — human OTP, run in user's own terminal** | Spending policies are **mainnet-only**. Testnet chains return an error if you try to set a policy. ## Prerequisites ```bash # Confirm session is good circle wallet status # Get the wallet address circle wallet list --chain BASE --type agent --output json ``` If `circle wallet status` errors with "Not logged in" or "Terms acceptance is required", hand off to the `use-agent-wallet` skill — it covers install, terms, login, and wallet creation. ## Viewing current limits (in-agent, no OTP) ```bash circle wallet limit --address <addr> --chain BASE --output json ``` Shows the current per-tx, daily, weekly, and monthly USDC caps (`null` for any unset tier). Safe to call freely — read-only, no money moves, no OTP. ## Setting or resetting limits (interactive terminal — handoff to user) `circle wallet limit set` and `circle wallet limit reset` are **interactive**. They send a 6-digit OTP to the user's email mid-execution and wait for the code at the CLI's stdin prompt. **OTPs are password-equivalent. The agent must NOT receive, store, or relay the OTP.** The agent's job here is to hand the user a verbatim command to run in their own terminal, then wait for them to report back. ### Step 1 — Confirm values with the user Limits must be **monotonic**: `per-tx ≤ daily ≤ weekly ≤ monthly`. A typical conservative configuration: | Tier | Suggested USDC value | |---|---| | `--per-tx` | `1` | | `--daily` | `5` | | `--weekly` | `20` | | `--monthly` | `50` | Adjust based on the user's stated tolerance. Get explicit yes before generating the command. ### Step 2 — Hand the user the command Tell the user: > Setting spending limits requires an OTP I shouldn't see. Please run this command in your own terminal — the CLI will email you a 6-digit code; enter it at the prompt. Let me know once it completes. > > ```bash > circle wallet limit set \ > --address <addr> --chain BASE \ > --policy-type stablecoin \ > --per-tx 1 --daily 5 --weekly 20 --monthly 50 > ``` For reset, the verbatim command is: ```bash circle wallet limit reset --address <addr> --chain BASE --yes ``` Omit `--yes` if you want the user to see a confirmation prompt before the OTP is sent. ### Step 3 — Verify after the user reports done ```bash circle wallet limit --address <addr> --chain BASE --output json ``` Confirms the new caps. Surface them to the user. ## Troubleshooting | Symptom | Cause | Fix | |---|---|---| | User received multiple OTP emails | Command was re-run while a previous invocation was still waiting | Tell the user to use the **most recent** OTP only. Earlier ones are invalidated. | | OTP rejected with "prefix mismatch" | User entered an OTP from a previous request | Restart — each `set` / `reset` invocation has a fresh prefix. | | `Spending policies are mainnet-only` | Tried to set a policy on a testnet chain | Re-run with a mainnet `--chain` value (`BASE`, `MATIC`, etc.). | | `Limits must be monotonic` | per-tx > daily, daily > weekly, etc. | Re-check the values. `per-tx ≤ daily ≤ weekly ≤ monthly`. | | `Only agent wallets support spending policies` | Tried on a local (non-agent) wallet | Spending policies apply only to agent wallets created via `circle wallet create`. | ## Rules - NEVER ask the user to share the OTP with you. The OTP must go directly from the user's email to the CLI prompt in their own terminal — it should never touch the agent's context, tool calls, or any local file. - If email- or messaging-integration tools are connected (e.g., Gmail or Slack via MCP), the OTP can also be fetched through them — note the option to the user; how to handle it is their call. - ALWAYS hand the user a verbatim `circle wallet limit set` / `reset` command for them to run in their own terminal. Do NOT try to orchestrate the OTP from chat. - ALWAYS confirm proposed limit values with the user before generating the command. Show the per-tx / daily / weekly / monthly numbers explicitly and wait for explicit yes. - Spending policies are mainnet-only — testnet chains are rejected. Don't quote testnet examples. - Only agent wallets support spending policies, not local wallets. - After the user reports the change is done, run `circle wallet limit --address <addr> --chain BASE --output json` so the user sees the new caps. ## Reference Links - Full policy walkthrough (extended examples, edge cases): https://agents.circle.com/skills/wallet-policy.md - Circle Developer Docs: https://developers.circle.com/llms.txt ## Alternatives Trigger the `use-agent-wallet` skill instead when: - The user has not yet logged in to the CLI or accepted Terms. - `circle wallet status` errors with "Not logged in" or "Terms acceptance is required". - The user is asking about wallet setup, login, or session state — not policy. Trigger the `pay-via-agent-wallet` skill instead when: - The user wants to call, pay for, or use a paid x402 service. - Limits are already set and the user is ready to spend. Trigger the `fund-agent-wallet` skill instead when: - The user wants to add USDC before setting limits. - The user mentions deposit, fiat on-ramp, fiat purchase, QR-code transfer, or Gateway deposit. --- DISCLAIMER: This skill is provided "as is" without warranties, is subject to the [Circle Developer Terms](https://console.circle.com/legal/developer-terms), and output generated may contain errors and/or include fee configuration options (including fees directed to Circle); additional details are in the repository [README](https://github.com/circlefin/skills/blob/master/README.md).
Related in AI Agents
skill-development
IncludedComprehensive meta-skill for creating, managing, validating, auditing, and distributing Claude Code skills and slash commands (unified in v2.1.3+). Provides skill templates, creation workflows, validation patterns, audit checklists, naming conventions, YAML frontmatter guidance, progressive disclosure examples, and best practices lookup. Use when creating new skills, validating existing skills, auditing skill quality, understanding skill architecture, needing skill templates, learning about YAML frontmatter requirements, progressive disclosure patterns, tool restrictions (allowed-tools), skill composition, skill naming conventions, troubleshooting skill activation issues, creating custom slash commands, configuring command frontmatter, using command arguments ($ARGUMENTS, $1, $2), bash execution in commands, file references in commands, command namespacing, plugin commands, MCP slash commands, Skill tool configuration, or deciding between skills vs slash commands. Delegates to docs-management skill for official documentation.
reprompter
IncludedTransform messy prompts into well-structured, effective prompts — single or multi-agent. Use when: "reprompt", "reprompt this", "clean up this prompt", "structure my prompt", rough text needing XML tags and best practices, "reprompter teams", "repromptception", "run with quality", "smart run", "smart agents", multi-agent tasks, audits, parallel work, anything going to agent teams. Don't use when: simple Q&A, pure chat, immediate execution-only tasks. See "Don't Use When" section for details. Outputs: Structured XML/Markdown prompt, quality score (before/after), optional team brief + per-agent sub-prompts, agent team output files. Success criteria: Single mode quality score ≥ 7/10; Repromptception per-agent prompt quality score 8+/10; all required sections present, actionable and specific.
adaptive-compaction
IncludedAdaptive add-on policy and recovery layer that decides WHEN to compact, prune, snapshot, or fork -- replacing fixed-percent auto-compaction across Claude Code, Codex, and MCP-capable hosts. Trigger on auto-compact timing or damage: "when should I compact", "is it safe to compact now or start a fresh session", "auto-compact fires too early/mid-task", "switching to an unrelated task but the window still has space", "context rot", "answers get worse the longer the session runs", "the agent forgot the plan or my decisions after it summarized", "add a layer on top that manages context without changing the agent", raising autoCompactWindow to give the policy room, or installing/tuning a cross-tool compaction policy or PreCompact hook -- even when "compaction" is never said but the problem is context-window pressure or post-summarization memory loss. Do NOT use to summarize a conversation, build RAG, write a summarization prompt (decides WHEN not HOW), or answer max-context-length trivia.
agent-skill-creator
IncludedCreate cross-platform agent skills from workflow descriptions. Activates when users ask to create an agent, automate a repetitive workflow, create a custom skill, or need advanced agent creation. Triggers on phrases like create agent for, automate workflow, create skill for, every day I have to, daily I need to, turn process into agent, need to automate, create a cross-platform skill, validate this skill, export this skill, migrate this skill. Supports single skills, multi-agent suites, transcript processing, template-based creation, interactive configuration, cross-platform export, and spec validation.
llm-wiki
IncludedUse when building or maintaining a persistent personal knowledge base (second brain) in Obsidian where an LLM incrementally ingests sources, updates entity/concept pages, maintains cross-references, and keeps a synthesis current. Triggers include "second brain", "Obsidian wiki", "personal knowledge management", "ingest this paper/article/book", "build a research wiki", "compound knowledge", "Memex", or whenever the user wants knowledge to accumulate across sessions instead of being re-derived by RAG on every query.
skill-master
IncludedAgent Skills authoring, evaluation, and optimization. Create, edit, validate, benchmark, and improve skills following the agentskills.io specification. Use when designing SKILL.md files, structuring skill folders (references, scripts, assets), ingesting external documentation into skills, running trigger evals, benchmarking skill quality, optimizing descriptions, or performing blind A/B comparisons. Keywords: agentskills.io, SKILL.md, skill authoring, eval, benchmark, trigger optimization.