stripe-directory
Use when the user wants to find businesses, software, service providers, or partners for a specific industry, workflow, pain point, capability, or job to be done. Also use when the agent needs to programmatically purchase or consume a service. Use Stripe Directory to build a short relevant shortlist, even if the user does not mention Stripe Directory explicitly.
What this skill does
## Stripe Directory Search
Turn a vague market need into a short, relevant shortlist with `stripe directory search`. Use this even when the user never says “Stripe Directory” — any request to find vendors, tools, partners, or providers for a vertical, workflow, pain point, or job-to-be-done.
Most requests are **discovery** — find and compare services. That is the core job below. Some services are also **MPP-supported** (MPP = Machine Payment Protocol), meaning you (the agent) can pay their HTTP 402 (Payment Required) endpoint and consume them directly. When the user actually wants to *use or buy* a service, present those results and offer to purchase — see “Purchasing” at the end.
## Process
1. **Clarify only what’s missing**: buyer/vertical, job-to-be-done, must-have capability, geography (only if it matters).
1. **Search iteratively**: `stripe directory search "<query>" --format json`
- Short noun phrases, one angle per query; run 1-3, then broaden/narrow on results.
- Angles to cover: vertical → workflow → pain point → adjacent. Two examples:
- services/trades: vertical (`electrician software`, `electrical contractor`) → workflow (`field service management`, `dispatch invoicing estimates`) → pain point (`job scheduling`, `quote automation`) → adjacent (`home services automation`, `contractor crm`).
- SaaS/software: vertical (`b2b saas billing`, `developer tools`) → workflow (`subscription management`, `usage-based metering`) → pain point (`failed payment recovery`, `revenue recognition`) → adjacent (`analytics dashboards`, `customer onboarding`).
- Hard constraints → filters: `--countries-supported=US`, `--has-stripe-app=true`, `--link-supported=true`, `--stripe-projects-supported=true`.
- If the user wants to *use/buy* a service, also pass `--mpp-supported` in at least one search to find results you can pay for programmatically.
- Sparse niche? Raise `--limit` and try the next `--page` before concluding it’s empty.
1. **Dedupe & score** using `display_name`, `description`, `url`, `username` as evidence.
- Prefer results whose description/site clearly match the target workflow.
- Prefer more trust signals over fewer: Projects provider, Link enabled, Marketplace app, Stripe Verified. For buy/use intent, also prefer MPP-supported results.
- Thin description but strong brand/domain match → keep in a weaker bucket, don’t discard.
1. **Return a shortlist, not a dump** — 5-10 strong matches, grouped:
- **direct** / **adjacent** / **needs manual review**
- Each entry: name · why it matched · URL (· which query surfaced it, when useful).
- Projects providers: offer the follow-up. The JSON gives the exact commands under each result’s `projects.catalog_command` / `projects.install_command` (`stripe projects catalog <provider>`, `stripe projects add <provider>`).
- MPP-supported results: note they’re purchasable and include `mpp.slug` / `mpp.url`.
1. **Be honest about weak results** — if sparse or generic, say so and adjust: broaden, narrow, or try synonyms rather than padding with noise.
Always report the exact queries (and filters) you ran so the user can keep iterating.
## Purchasing (only when the user wants to buy or consume a service)
MPP-supported results are payable directly. Don’t drive to purchase unprompted. When the user wants to buy, **present the full menu of payment methods and ask which they’d like to use** before doing anything:
> "Which payment method would you like to use?
>
> - **Link CLI** — Stripe-native, test mode available (recommended)
- **Tempo** — crypto wallet
- **Privy Agent Wallet CLI** — crypto wallet
- **mppx** — debug-only fallback"
Once the user picks, silently run `which <tool> 2>/dev/null` to check if it’s installed. If not installed, offer to install it (for example, `npm i -g @stripe/link-cli` for Link CLI) and wait for confirmation before proceeding.
**Always show the price and get explicit user approval before any money moves**; prefer a no-charge test path first.
Short version:
1. Resolve the real callable endpoint from the result’s `mpp.slug` / `mpp.url`. `mpp.url` is often the mpp.dev landing form (`https://mpp.dev/services#<slug>`) — resolve the raw endpoint on [mpp.dev](https://mpp.dev) if so. Read the HTTP 402 challenge to confirm the amount: `curl -s -D - -o /dev/null <endpoint_url>` (look for `WWW-Authenticate`).
1. Use the payer the user selected.
- **`link-cli`** (Stripe-native Shared Payment Token, has a test mode, no crypto wallet, US Link accounts only; `npm i -g @stripe/link-cli`): `auth login` → `mpp decode --challenge "<value>"` (get `network_id`) → `spend-request create --credential-type shared_payment_token --network-id <id> --amount <cents ≤50000> --context "<100+ chars>" --request-approval` (blocks for approval) → `mpp pay <endpoint_url> --spend-request-id <approved_id>`.
- **Tempo**: `tempo wallet login` / `services` / `request`.
- **Privy**: `@privy-io/agent-wallet-cli`.
- **mppx**: debug-only fallback.
Never invent results or skip the price/approval gate.
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.