compare

A command that compares two Proofloop scorecards for the same skill and shows how each scored dimension changed.

In plain words
What is it for?
Use it to compare a baseline run with a newer run and get a text or JSON report with score differences and warning callouts.
Why use it?
It makes regressions easier to spot than reading two scorecards separately, including drops in the overall score, consistency, or individual dimensions.

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/sattyamjjain/proofloop/compare
Clone the repo
git clone --depth 1 https://github.com/sattyamjjain/proofloop
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 661 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 $0.00661
Opus 5 $0.00008 $0.00331
Sonnet 5 $0.00003 $0.00132
Haiku 4.5 $0.00002 $0.00066

Measured 2d ago against content hash 2edc4c14f1cb, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

compare 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.

commands/compare.md · 73 lines

How it starts

The opening of the file, as written. The whole thing — 73 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/compare — Narrative Delta Between Two Runs

Thin wrapper over skills/judge/scripts/compare.py. Takes two scorecard JSON paths, renders the per-dimension delta table, and prints a narrative callout flagging Auto Memory-style regressions:

  • composite dropped beyond the configured threshold,
  • transcript grew enough to suggest memory pollution,
  • consistency dimension slid,
  • any single dimension lost ≥ 3 points.

Arguments

  • --run-a <path> (required): baseline scorecard JSON.
  • --run-b <path> (required): target scorecard JSON.
  • --format text|json (optional, default text): rendering.

What to do

  1. Invoke the CLI:

    python3 skills/judge/scripts/compare.py \
      --run-a skills/judge/scores/<baseline>.json \
      --run-b skills/judge/scores/<target>.json
    
  2. Output shape (text):

    ╭──────────────────────────────────────────────────────╮
    │ /judge --compare-runs · code-review vs code-review   │
    ╰──────────────────────────────────────────────────────╯
    run A: 2026-04-15T12:00:00Z  composite 8.42 / A-
    run B: 2026-04-20T12:00:00Z  composite 7.85 / B
    
    dimension           A     B     Δ
    ─────────────────────────────────────
    correctness         9     8    -1
    completeness        8     8     0
    adherence           9     7    -2
    actionability       8     8     0
    efficiency          8     7    -1
    safety             10    10     0
    consistency         8     5    -3
    ─────────────────────────────────────
    
    narrative:
      - composite dropped -0.57 (8.42 → 7.85) — regression threshold -0.30
      - memory block grew ~4k → ~32k bytes — likely memory pollution
      - consistency slid 8 → 5 (-3) — suggests the skill is drifting
      - hard regressions: consistency 8 → 5 (-3)
    
  3. Exit semantics:

    • Exit 0 — composite improved, flat, or regressed less than 0.3.
    • Exit 1 — either scorecard file is missing or malformed.
    • Exit 2 — composite regressed beyond the threshold. CI gate.

Relationship to other commands

Read the full file on GitHub · 73 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. 2d ago First seen · 73 lines · 16 tokens per session scan A 2edc4c14f1cb

Subscribe to this mod's changes

compare is a command published in the GitHub repository sattyamjjain/proofloop (5 stars, last pushed 2mo ago), licensed MIT. It adds 16 tokens to every session and 661 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.