ijfw-assumptions-analyzer

ijfw-assumptions-analyzer is an agent for Claude Code from FerroxLabs/ijfw. It costs 34 tokens per session (1,679 once invoked), scanned A, original, MIT.

A review step that finds hidden assumptions in a brief and its plan before work begins. An assumption is something the plan relies on even though the requirements do not guarantee it.

In plain words
What is it for?
Auditing plans for software projects, campaigns, designs, launches, and other work before execution, then recording each assumption and a cheap way to check it.
Why use it?
It exposes gaps that could cause a project to fail later, such as assuming data fits in memory or that an external service returns a certain format.

Agent for Claude Code

Written for Claude Code: allowed-tools in frontmatter. Also seen: model in frontmatter; mentions subagents.

Part of the ijfw plugin — 34 skills, 22 commands, 37 agents, 6 hooks shipped together

Good fit Auditing plans for software projects, campaigns, designs, launches, and other work before execution, then recording each assumption and a cheap way to check it.

Compare 6 agents from other repositories ↓
Install with agentmods
npx agentmods add agents/ferroxlabs/ijfw/ijfw-assumptions-analyzer
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.

Clone the repo
git clone --depth 1 https://github.com/FerroxLabs/ijfw

Made for: Claude Code.

Or install ijfw, the plugin that ships this one along with the rest of its 34 skills, 22 commands, 37 agents, 6 hooks.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for ijfw-assumptions-analyzer

README.md
[![agentmods](https://agentmods.dev/badge/agents/ferroxlabs/ijfw/ijfw-assumptions-analyzer.svg)](https://agentmods.dev/agents/ferroxlabs/ijfw/ijfw-assumptions-analyzer)
Your own site
<a href="https://agentmods.dev/agents/ferroxlabs/ijfw/ijfw-assumptions-analyzer"><img src="https://agentmods.dev/badge/agents/ferroxlabs/ijfw/ijfw-assumptions-analyzer.svg" alt="Measured on agentmods" height="20"></a>
Per session 34 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,679 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00034 $0.01679
Opus 5 $0.00017 $0.00839
Sonnet 5 $0.00007 $0.00336
Haiku 4.5 $0.00003 $0.00168

Measured 8d ago against content hash dbd230acf12b, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-08, from the pricing page.

Security

Grade A, and why

ijfw-assumptions-analyzer 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 8d 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.

claude/agents/ijfw-assumptions-analyzer.md · 180 lines

How it starts

The opening of the file, as written. The whole thing — 180 lines — stays where its author put it; the contents beside it link to each section on GitHub.

ijfw-assumptions-analyzer -- hidden-assumption surfacing

You are an IJFW assumption-surfacing subagent. You read a brief and a plan, then produce ASSUMPTIONS.md: a structured ledger of everything the plan takes for granted that the brief / spec does NOT guarantee. The goal is to make the implicit explicit BEFORE execution begins, so the orchestrator can validate, narrow, or escalate the risky ones.

Domain-agnostic. Works for software, books, campaigns, designs, launches -- anywhere a brief is handed to a plan and the plan fills in gaps.

ROLE

Hidden-assumption auditor. Most plans fail not because the work is hard but because the plan quietly assumes something the brief never promised: "the data fits in memory", "the reader knows X", "the user is logged in", "the API returns JSON". When that quiet assumption is wrong, execution escalates mid-stream. This agent surfaces those gaps at plan-time.

You do NOT grade the plan, propose alternatives, or rewrite. You only name the assumptions, classify them, and suggest the cheapest validation step.

PROCESS

  1. Locate brief + plan.

    • Default brief: .ijfw/memory/brief.md.
    • Default plan: .ijfw/memory/plan.md.
    • Fallback: if invoked with phase input, look under .planning/<milestone>/<phase>/SPEC.md (brief) and .planning/<milestone>/<phase>/PLAN.md (plan).
    • If either source is missing, emit a MISSING_INPUT finding and stop -- do not invent content.
  2. Read both fully. Use Read. For long files, read in chunks; do not skim. The whole job is catching what was glossed over.

  3. Diff the surface. For every concrete claim in the plan, ask: "Did the brief actually guarantee this?" Three kinds of gap matter:

    • Hard assumption -- if false, the plan WILL fail (data shape, auth model, runtime availability, ordering guarantee, audience literacy, distribution channel access).
    • Soft assumption -- if false, quality degrades but plan still ships (perf target, tone, edge-case coverage, polish level).
    • Implicit dependency -- unstated reliance on environment, config, prior work, third-party service, reader prerequisite, or data the brief doesn't promise to provide.

Read the full file on GitHub · 180 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. 8d ago First seen · 180 lines · 34 tokens per session scan A dbd230acf12b

Subscribe to this mod's changes

ijfw-assumptions-analyzer is an agent published in the GitHub repository FerroxLabs/ijfw (210 stars, last pushed yesterday), licensed MIT. It adds 34 tokens to every session and 1,679 once invoked, about $0.0002 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 2026-08-30.