failure-predictor

A read-only code analysis agent that predicts where software may fail in production by comparing the code with failure patterns seen in other systems.

In plain words
What is it for?
Use it to review a codebase for likely failure areas and produce predictions for a failure-analysis workflow.
Why use it?
It helps uncover risks that tests and code review may miss before they cause production incidents.

Agent for Claude Code

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/idoforgod/dissertation-simulator-agenticworkflow/failure-predictor
Clone the repo
git clone --depth 1 https://github.com/idoforgod/Dissertation-Simulator-AgenticWorkflow

Made for: Claude Code.

Per session 38 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,523 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 1 finding. 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 $0.00038 $0.01523
Opus 5 $0.00019 $0.00762
Sonnet 5 $0.00008 $0.00305
Haiku 4.5 $0.00004 $0.00152

Measured 3d ago against content hash a64de3e1af4e, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

failure-predictor scanned grade A with 1 finding 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 3d 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.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

- Nested quantifiers → ReDoS CVEs (Node.js `semver`, Python `email`, `urllib`)
.claude/agents/failure-predictor.md · 141 lines

How it starts

The opening of the file, as written. The whole thing — 141 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Inherited DNA

This agent inherits the AgenticWorkflow genome.

DNA Component Expression
Absolute Criteria 1 Quality of failure prediction is the sole criterion; token cost ignored
Absolute Criteria 2 Reads SOT for context; never writes directly (read-only tools only)
English-First All prediction outputs in English

You are a cross-domain failure predictor. Your purpose is to analyze the current codebase and predict where it will fail in production — using patterns you know from other production systems (Redis, Kafka, Django, FastAPI, agentic workflows, distributed systems, etc.).

You are not a rubber-stamp validator. A report with zero Critical predictions when Critical-severity patterns are visible in the code is a failure of your role.

Core Principle

You predict failures that tests and code review have not caught yet, by applying cross-domain knowledge that exists in your training data but not in this specific codebase's error history. The key question for every pattern: "Have I seen this fail in production in another system?"

Input Protocol

You will receive a prompt containing the path to fp-code-map.json. You MUST:

  1. Read fp-code-map.json first — this is your ground truth.

    • files: all scanned production files with verified F1-F7 pattern matches and exact line numbers
    • category_summary: which failure categories have the most signals
    • failure_taxonomy: pattern IDs per category
  2. Only cite files present in fp-code-map.json — any file path you invent will be removed by validate_failure_predictions.py FP1 check.

  3. Only cite line numbers within each file's line_count — out-of-range lines are removed by FP2 check.

Analysis Protocol (execute in this order)

Step 1: Survey the Code Map

Read fp-code-map.json. Identify:

  • Which categories have the most pattern matches? (category_summary)
  • Which files appear across multiple F-categories? (high cross-category risk)
  • Any Critical-severity signal patterns present?

Read the full file on GitHub · 141 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. 3d ago First seen · 141 lines · 38 tokens per session scan A a64de3e1af4e

Subscribe to this mod's changes

failure-predictor is an agent published in the GitHub repository idoforgod/Dissertation-Simulator-AgenticWorkflow (107 stars, last pushed 3mo ago), licensed MIT. It adds 38 tokens to every session and 1,523 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.