fire-discover

fire-discover is a command for coding agents from ThierryN/fire-flow. It costs 7 tokens per session (5,159 once invoked), scanned A, original, MIT.

A command that reviews completed work to find repeated problem-solving patterns and suggest reusable skills or updates to existing ones.

In plain words
What is it for?
Use it to discover recurring techniques, review recent project records and code changes, create weekly summaries, and export findings as Markdown or JSON.
Why use it?
It helps turn lessons from past projects into documented, repeatable workflows instead of leaving them scattered across files and commits. It can analyze a recent phase, a week of work, or selected analysis depth.

Command

Part of the fire-flow plugin — 44 commands, 15 agents shipped together

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 commands/thierryn/fire-flow/fire-discover
Clone the repo
git clone --depth 1 https://github.com/ThierryN/fire-flow

Or install fire-flow, the plugin that ships this one along with the rest of its 44 commands, 15 agents.

Per session 7 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 5,159 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 $0.00007 $0.05159
Opus 5 $0.00003 $0.02580
Sonnet 5 $0.00001 $0.01032
Haiku 4.5 $0.00001 $0.00516

Measured 3d ago against content hash d3b520c70347, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

fire-discover 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 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.

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.

commands/fire-discover.md · 617 lines

How it starts

The opening of the file, as written. The whole thing — 617 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/fire-discover

AI-powered pattern discovery and skill suggestion engine


Purpose

Analyze completed work to detect recurring patterns, problem-solving approaches, and common solutions that could become reusable skills. Automatically suggests creating new skills or updating existing ones based on observed patterns. Helps compound knowledge across projects.


Arguments

Argument Required Description
--phase [N] No Analyze specific phase (default: most recent completed)
--weekly No Generate weekly pattern summary
--auto No Non-interactive mode, auto-defer all suggestions
--depth [level] No Analysis depth: shallow, normal, deep (default: normal)
--export [format] No Export findings as md or json

Process

Step 1: Gather Analysis Data

Collect data from completed work.

# Data sources for pattern detection
SUMMARIES=".planning/phases/*/RECORD.md"
COMMITS=$(git log --oneline --since="7 days ago")
CHANGED_FILES=$(git diff --stat HEAD~20..HEAD)
CODE_COMMENTS=$(grep -r "// tricky\|// hard\|// discovered\|// pattern" src/)

Sources analyzed:

  • RECORD.md files - Skills applied, problems solved, decisions made
  • Git commits - Commit messages indicating patterns or solutions
  • Code comments - Developer annotations about non-obvious solutions
  • File changes - Structural patterns in code organization
  • Test files - Testing patterns and edge cases handled

Step 2: Pattern Detection

Identify recurring patterns using multiple detection methods.

Detection Methods:

1. Code Pattern Analysis (via AST concepts)
   - Function signatures that repeat
   - Similar error handling approaches
   - Common utility patterns
   - Component structure templates

2. Problem-Solving Approaches (from commits)
   - "fix:" commits that address similar issues
   - Refactoring patterns that repeat
   - Performance optimization techniques
   - Security hardening approaches

3. Frequency Analysis
   - Code snippets that appear 3+ times
   - Similar file structures
   - Repeated import patterns
   - Common configuration shapes

Read the full file on GitHub · 617 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 · 617 lines · 7 tokens per session scan A d3b520c70347

Subscribe to this mod's changes

fire-discover is a command published in the GitHub repository ThierryN/fire-flow (77 stars, last pushed 21d ago), licensed MIT. It adds 7 tokens to every session and 5,159 once invoked, about $0.0000 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.