spike

A command that investigates exactly one technical assumption through a four-stage process: design, execution, written findings, and review. A spike is a small experiment used to answer whether an approach works, and TDD is not required here.

In plain words
What is it for?
Use it to run a named technical experiment, continue an interrupted experiment, choose a deeper executor, and produce validated notes and review artifacts.
Why use it?
It replaces guesswork with measured evidence and records a derived YES, NO, or UNCERTAIN result that can be resumed and checked.

Command

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/chris-dare-dev/agent-kit/spike
Clone the repo
git clone --depth 1 https://github.com/chris-dare-dev/agent-kit
Per session 136 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 4,925 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 $0.00136 $0.04925
Opus 5 $0.00068 $0.02462
Sonnet 5 $0.00027 $0.00985
Haiku 4.5 $0.00014 $0.00492

Measured yesterday against content hash 91ac118a85ce, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

spike 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 yesterday.

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.

data/commands/spike.md · 352 lines

How it starts

The opening of the file, as written. The whole thing — 352 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Spike Pipeline Orchestrator

A spike answers exactly one assumption with measured data. POC code is throwaway; the durable artifacts are note.md and review.md. In spike v2 the run is governed by a machine state file (state.json, owned by spike-checkpoint.py) — every phase transition is evidence-gated, the verdict is derived (spike-decide.py), and completion is never inferred from which files happen to exist. Read data/references/spike-protocol.md, spike-state-schema.md, spike-artifact-schema.md, and spike-agent-contract.md.

Arguments: $ARGUMENTS — parse as <spike-id> [--resume] [--executor deep] [--skip-review]

  • <spike-id> — required. Everything before the first --. Shape: ^[a-z][a-z0-9-]*-spike-[0-9]+$ (validated in Step 0). The id is <topic>-spike-<N>.
  • --resume — read state.json; continue from the phase it records. (Resume is state-driven regardless of this flag; it just skips the "fresh vs resume" prompt.)
  • --executor deep — run the executor at policy class deep-reasoning-high (dispatch model override — pass only the model token: fable (effort: high)) instead of its default class balanced-high. Roughly doubles cost. Legacy spelling --executor opus is accepted.
  • --skip-review — stop after Phase 3 (note.md); record the audited skip-review terminal (which still requires a canonical derived verdict). Use only for quick internal spikes.

If $ARGUMENTS is empty or has no spike-id, STOP and ask: "What is the spike id? (e.g., kiali-multicluster-spike-1)".


When to use / When NOT to use

Use /spike Use /milestone-pipeline Use neither
Answering ONE binary/three-way assumption Shipping code that implements a feature Exploration without a specific question
Validating a design decision before committing Full research → implement → critique → rectify Refactors, lint, doc-only changes
Time-boxed, throwaway POC Durable, reviewed, committed code Single-file fixes (use Edit directly)

Read the full file on GitHub · 352 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. yesterday First seen · 352 lines · 136 tokens per session scan A 91ac118a85ce

Subscribe to this mod's changes

spike is a command published in the GitHub repository chris-dare-dev/agent-kit (0 stars, last pushed 24d ago), licensed MIT. It adds 136 tokens to every session and 4,925 once invoked, about $0.0007 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.