Metric Validation Harness

What it does

When your agent uses it

• Ever someone proposes
• Asks is this metric any good
• Suspects a score tracks LOC
• Jumps between runs

Example prompts

• “is this metric any good”
• “/metric-validation-harness”

Requirements

• Python 3
• A Bash shell

What it can do on your machine

Read from SKILL.md and the folder at commit cf93c57. It shows what the files ask for, not the result of running them.

• Tool permissions

  Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

  From allowed-tools in the SKILL.md frontmatter.

• Runs code

  Ships 10 files in scripts/ (Python and Shell, from the files we listed), which the agent can run.

  Shell commands in SKILL.md call:

  • bash
  • python3

  From the folder's file list and the shell code blocks in SKILL.md.

• Network

  No URLs in SKILL.md.

  From URLs in SKILL.md, links to its own repository left out.

• Credentials

  Names no API keys, tokens, secrets or passwords.

  From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Metric Validation Harness loads about 1.7k tokens when it runs, and up to ~3.4k if it reads all its reference files. Until then it costs about 233 tokens; SKILL.md has 627 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~233

When it runs · the whole SKILL.md, loaded when a task matches

~1.7k

With references · SKILL.md plus every file in references/, read only if the agent opens them

~3.4k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.

SKILL.md

The full file from pproenca/dot-skills at commit cf93c57, republished under its MIT licence (© pproenca). 627 words, ~1,735 tokens.

name

metric-validation-harness

description

Empirically validates a software metric before trusting or optimizing it — point it at any candidate metric (a command that takes a path and prints one number) plus a corpus, and it runs experiments that try to falsify each property a good metric must have. Checks determinism (same input, same number across runs and hash seeds), invariance to cosmetic edits (also an anti-gaming probe), monotonicity under construct-increasing edits, discrimination, robustness on edge inputs, near-linear tractability, and construct validity (convergent, discriminant vs LOC, predictive AUC, lift over a baseline). Trigger whenever someone proposes, reviews, tunes, or ships a metric, score, or index, asks "is this metric any good", suspects a score tracks LOC or jumps between runs, or builds a deterministic optimization target. It is the empirical companion to the deterministic-metric-design skill and is read-only.

config.json / env  →  resolve metric_cmd, corpus, thresholds (env > config > bundled default)
        │
        ▼
   verify.sh ──► determinism ─ invariance ─ monotonicity ─ robustness ─ tractability ─ validity
        │            (each property check maps to a deterministic-metric-design category)
        ▼
   PASS / FAIL per property  →  exit 0 (all pass) or 1 (any group failed)

$ python3 mymetric.py path/to/file.py
42

# 1. Validate the bundled example metric (works with zero setup):
bash scripts/verify.sh

# 2. Validate YOUR metric — set metric_cmd in config.json, or override per-run:
METRIC_CMD="python3 /abs/path/mymetric.py" bash scripts/verify.sh

# 3. Prove the harness itself works (positive + negative cases):
bash scripts/selftest.sh

# 4. Sanity-check your adapter prints one number:
bash scripts/run-metric.sh path/to/file.py

© pproenca, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 27 other files (scripts, references) in skills/.experimental/metric-validation-harness of pproenca/dot-skills.

• SKILL.md
• .gitignore
• config.json
• gotchas.md
• metadata.json
• references/workflow.md
• scripts/check-determinism.sh
• scripts/check-invariance.sh
• scripts/check-monotonicity.sh
• scripts/check-robustness.sh
• scripts/check-tractability.py
• scripts/check-validity.py
• scripts/examples/metric_ast_nodes.py
• scripts/examples/metric_loc.py
• scripts/fixtures/corpus.csv
• scripts/fixtures/corpus/f1_dense.py
• … and 12 more

Open the folder on GitHubat commit cf93c57