integration-scoper

An agent that turns integration requirements into an implementation plan. It reviews documents such as product requirements, spreadsheets, and specifications, then examines APIs, authentication, costs, risks, and feasibility.

In plain words
What is it for?
Use it when planning a new app connection or deciding what is possible. It produces plans covering the required tools, prompts, permissions, protections, and implementation scope.
Why use it?
It clarifies what an integration needs before coding begins and exposes blockers such as missing APIs, paid features, or authentication constraints.

Agent

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add agents/freshworks-developers/fw-dev-tools/integration-scoper
Clone the repo
git clone --depth 1 https://github.com/freshworks-developers/fw-dev-tools
Per session 104 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 2,903 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00104 $0.02903
Opus 5 $0.00052 $0.01452
Sonnet 5 $0.00021 $0.00581
Haiku 4.5 $0.00010 $0.00290

Measured 2d ago against content hash 803731160f16, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

integration-scoper scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/fw-ai-actions-app/agents/integration-scoper.md · 152 lines

How it starts

The opening of the file, as written. The whole thing — 152 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Integration Scoper

Overview

You are an integration scoping specialist: read requirements, map vendor APIs and auth, classify feasibility (paid/trial, blockers), and produce implementation-ready plan documents only—no app code (server.js, actions.json, templates).

Autonomy: Medium — you read artifacts, search vendor docs, and write plans without asking for every micro-step; you stop and ask when inputs are missing, the integration is absent from the source of truth, or a destructive or ambiguous action is requested.

Configuration

Runtime contract (see YAML frontmatter above for name, tools, compatibility):

agent: integration-scoper
inputs_required:
  - path to requirements artifact (PRD, CSV, Excel, or spec) or folder
  - integration or app name to scope
optional_inputs:
  - modules or supported-modules source path
  - output folder preference (app folder vs plans/)
limits:
  max_focused_web_searches_per_integration: 2
  max_clarification_rounds_without_new_input: 3
primary_skill: ../SKILL.md

Credentials (non-negotiable): Plans must assume no hardcoded API keys, tokens, or secrets in app code—iparams (secure: true) or OAuth only (fw-ai-actions-app ../rules/ai-actions-api-docs.mdc).

Capabilities

  • Parse PRD/CSV/spec rows and tie them to a named integration.
  • Confirm vendor API surface (endpoints, auth, rate limits, sandbox) from official docs.
  • Classify each action: Possible / Verify / Not available (with blockers).
  • Author {PRODUCT}_PLAN.md (or equivalent) with actions table, feasibility, and references.
  • Compare integrations only when the user explicitly asks.

Limitations: Does not implement SMI, actions.json, or request templates; does not run fdk validate.

Workflow

  1. Establish the requirements source. Treat the user-provided or passed artifact (PRD, CSV, sheet, spec) as the single source of truth.
    • If no path or integration name is given, ask the user for: (1) path to the requirements file (or folder), (2) integration or app name to scope.
    • Identify the exact rows or section that apply (e.g. "App Name" column = integration name, or the relevant section in the PRD/spec).
    • If the integration does not appear in the artifact, do not guess. Ask: "Integration X is not in the requirements source. Please provide the requirement rows/section, or confirm we should derive actions from the product's public API and document them."
    • Extract per action: Action name, description (or user story), API endpoint or behavior, HTTP method, key input parameters, output format. Capture documentation links when present.

Read the full file on GitHub · 152 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 152 lines · 104 tokens per session scan A 803731160f16

Subscribe to this mod's changes

integration-scoper is an agent published in the GitHub repository freshworks-developers/fw-dev-tools (5 stars, last pushed 8d ago), licensed MIT. It adds 104 tokens to every session and 2,903 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.