replay

A command that rebuilds the history of a Wheat research sprint from earlier versions of its claims file. It recompiles each version and creates a timeline showing how the sprint changed.

In plain words
What is it for?
Use it to find phase transitions, the first ready compilation, conflict peaks, evidence improvements, and claims that were later replaced.
Why use it?
It makes changes in evidence, conflicts, phases, and conclusions visible instead of showing only the current state.

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/grainulation/wheat/replay
Clone the repo
git clone --depth 1 https://github.com/grainulation/wheat
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 497 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 $0.00000 $0.00497
Opus 5 $0.00000 $0.00249
Sonnet 5 $0.00000 $0.00099
Haiku 4.5 $0.00000 $0.00050

Measured 2d ago against content hash 7d74fc4304df, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

replay 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 2d 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.

templates/commands/replay.md · 70 lines

What it actually says

/replay — Time-Travel Through Sprint Evolution

You are reconstructing the historical evolution of this sprint by recompiling every version of claims.json from git history.

Process

  1. Get the git history of claims.json:

    git log --oneline claims.json
    

    This gives every commit that touched claims.json — the sprint event log.

  2. Extract each historical version: For each commit hash:

    git show <hash>:claims.json > /tmp/wheat-replay-<N>.json
    
  3. Recompile each version with the current compiler:

    npx @grainulation/wheat compile --input /tmp/wheat-replay-<N>.json --output /tmp/wheat-comp-<N>.json
    
  4. Compute deltas between consecutive compilations:

    npx @grainulation/wheat compile --diff /tmp/wheat-comp-<N-1>.json /tmp/wheat-comp-<N>.json
    
  5. Identify interesting moments in each delta:

    • Phase transitions (define -> research -> prototype -> evaluate)
    • First time compilation went "ready"
    • Peak conflict count
    • Evidence tier jumps (topic going web -> tested)
    • Claims added then superseded (the sprint changed its mind)
  6. Generate replay HTML: Create output/replay.html — a self-contained timeline visualization using a dark scroll-snap template. Include:

    • Frame-by-frame scrubbing (each commit = one frame)
    • Highlighted pivotal moments
    • Coverage evolution chart
    • Summary statistics per frame
  7. Print a text summary to the terminal with the key narrative moments.

Git commit

Commit: wheat: /replay — generated sprint timeline (N frames)

Tell the user

  • How many frames (commits) were found
  • The most interesting moments
  • Point them to output/replay.html for the full interactive timeline
  • Suggest: /handoff to package this narrative for a successor

Cleanup

Remove temporary files from /tmp after generating the output:

rm -f /tmp/wheat-replay-*.json /tmp/wheat-comp-*.json

$ARGUMENTS

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. 2d ago First seen · 70 lines · 0 tokens per session scan A 7d74fc4304df

Subscribe to this mod's changes

replay is a command published in the GitHub repository grainulation/wheat (20 stars, last pushed 4mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 497 tokens. 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.