blast-radius

blast-radius is a command for coding agents from andyzengmath/soliton. It costs 16 tokens per session (711 once invoked), scanned A, original, MIT.

A command that estimates a file's blast radius: how many other files refer to it and whether it matches sensitive file paths. A sensitive path is a location that may require extra care, such as configuration or security-related code.

In plain words
What is it for?
Run it on a repository file to count references and check whether the file falls under configured sensitive-path rules.
Why use it?
It helps assess the likely impact and risk of changing a file before editing it.

Command

Part of the soliton plugin — 1 skill, 3 commands, 13 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/andyzengmath/soliton/blast-radius
Clone the repo
git clone --depth 1 https://github.com/andyzengmath/soliton

Or install soliton, the plugin that ships this one along with the rest of its 1 skill, 3 commands, 13 agents.

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 blast-radius

README.md
[![agentmods](https://agentmods.dev/badge/commands/andyzengmath/soliton/blast-radius.svg)](https://agentmods.dev/commands/andyzengmath/soliton/blast-radius)
Your own site
<a href="https://agentmods.dev/commands/andyzengmath/soliton/blast-radius"><img src="https://agentmods.dev/badge/commands/andyzengmath/soliton/blast-radius.svg" alt="Measured on agentmods" height="20"></a>
Per session 16 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 711 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 $0.00016 $0.00711
Opus 5 $0.00008 $0.00356
Sonnet 5 $0.00003 $0.00142
Haiku 4.5 $0.00002 $0.00071

Measured 3d ago against content hash 1e1b6e5caab2, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

blast-radius 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/blast-radius.md · 54 lines

How it starts

The opening of the file, as written. The whole thing — 54 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/blast-radius

Compute the blast radius (number of files that import or reference a target file, plus sensitive-paths hit) using Soliton's risk-scorer heuristic — the same grep-based computation the risk-scorer agent uses today (agents/risk-scorer.md Factor 1).

This is a degraded-mode standalone surface of Soliton's existing logic. When the graph plugin lands (POST_V2_FOLLOWUPS §B1), this command will switch to graph-cli's info + dependency-edges queries; the user-facing contract stays the same.

Argument

  • <file> — path to a file in the current repository (relative or absolute).

Behavior

  1. Validate input: if <file> is missing or doesn't exist, output Error: file not found: <file> and STOP.

  2. Compute import-count via grep heuristic (same as agents/risk-scorer.md Factor 1):

    • Extract the file's basename without extension (e.g. auth-utils.tsauth-utils)
    • Run git grep -l "<basename>" to count files that reference it
    • Strip the input file itself from the count
  3. Compute sensitive-paths hit (same as risk-scorer Factor 3):

    • Apply each pattern from rules/sensitive-paths.md (and any config.sensitivePaths overrides from .claude/soliton.local.md) to the input file path
    • Mark hit if any pattern matches
    • Note: do NOT inline the pattern list in this command's output — read the canonical source so future updates to rules/sensitive-paths.md propagate automatically.
  4. Output a compact summary:

Blast Radius for <file>

Importers: <N> files (grep heuristic)
Sensitive: <hit | clean>
Backing source: grep (graph plugin pending §B1; will upgrade when graph-cli ships)

Top 5 importers (alphabetical):
- path/a.ts
- path/b.ts
- ...

If the import-count is 0, note explicitly: "0 importers found via grep — file may be entry-point, deleted-only PR target, or grep-heuristic miss. Risk-scorer's full analysis runs the same query against git diff context."

Non-goals

  • This is NOT the full risk score. The risk-scorer agent computes 6 factors weighted to a 0-100 score; /blast-radius exposes only Factor 1 + Factor 3 in isolation for ad-hoc queries.
  • This is NOT graph-aware. When the graph plugin lands, the same surface upgrades to use real call/import edges. Until then, false-positives are possible (any file containing the basename string matches, including comments and unrelated symbols).

Read the full file on GitHub · 54 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 · 54 lines · 16 tokens per session scan A 1e1b6e5caab2

Subscribe to this mod's changes

blast-radius is a command published in the GitHub repository andyzengmath/soliton (1 stars, last pushed 2mo ago), licensed MIT. It adds 16 tokens to every session and 711 once invoked, about $0.0001 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.

Related

Other commands, from other repositories

harness-build

Full build pipeline — evidence gate, product gate, COGS gate, PRD writing, OKR design, sprint planning, design system setup, spec compliance review (--step spec-review), quality gate (--step quality-gate), progress tracking, and scope guard. Use when committing to building a product or agent, running individual gate…

kimsanguine/hplan · 81 tokens

harness-operate

Full production operations workflow — KPI dashboard, reliability scan, cost review, improvement planning, knowledge extraction, decision pattern matching, and TK-to-instruction conversion. Use when an agent is live and needs a weekly/monthly operational review, or when running any individual operations step. Use…

kimsanguine/hplan · 73 tokens

harness-verify

Verify task/feature completion against CONDITIONALGO conditions in STATE.md. Syncs condition status (❌→✅) before any 'done' declaration. Use when marking a task complete, closing a sprint, or claiming a condition is met. Use --spec to run full Spec Compliance Review instead of STATE.md condition check.

kimsanguine/hplan · 65 tokens

openehr-explain

One-stop router that explains or looks up any openEHR thing — auto-detects an archetype, a template, an RM/AM/BASE type, an RM structural concept, an ADL idiom, an AQL query or keyword, or a terminology code (replaces /archetype-explain, /template-explain, /type-spec, /rm-structure, /adl-idiom, /terminology).

Cadasto/openehr-assistant-plugin · 95 tokens

cogs-sentinel

Use when AI 기능의 단가·사용량·가격으로 p50/p90 월간 마진을 결정해야 합니다. hplan/cogs-sentinel wrapper는 lognormal sampler와 free-user abuse blend로 GREEN/CONDITIONALGO/RED를 판정합니다.

kimsanguine/hplan · 56 tokens

harness

Run the full Auto-Harness orchestration flow. /auto-harness:harness.

redker56/auto-harness · 25 tokens