Agent skill

Allowed Diffs

by lidofinance in lidofinance/diffyscan

Adds, reviews or tightens Diffyscan alloweddiffs rules for explained source or bytecode differences, including replacing any wildcards with granular rules.

MITAuto-check passedBackend & APIs

Install Allowed Diffs

skills CLI
$ npx skills add lidofinance/diffyscan --skill allowed-diffs -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install lidofinance/diffyscan allowed-diffs --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/lidofinance/diffyscan.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.claude/skills/allowed-diffs .claude/skills/allowed-diffs && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
allowed-diffs
GitHub stars
142
Token cost
~1.2k tokens
SKILL.md length
592 words
Files
2 (incl. references)
Skills in repo
5
Repo updated
First seen
Licence
MIT

At a glance

Adds, reviews or tightens Diffyscan alloweddiffs rules for explained source or bytecode differences, including replacing any wildcards with granular rules.

  • Works in 3 steps: Establish evidence → Choose a rule → Verify the result
  • Accepting an expected diff
  • SKILL.md covers 1. Establish evidence, 2. Choose a rule and 3. Verify the result
  • Calls uv

What it does

Allowed Diffs is an agent skill from lidofinance/diffyscan. Adds, reviews or tightens Diffyscan alloweddiffs rules for explained source or bytecode differences, including replacing any wildcards with granular rules. Use when accepting an expected diff or narrowing an exception; use debug-diff first if its cause is unknown.

Its SKILL.md is about 1.2k tokens, which your agent loads only when the skill is triggered. The skill folder holds 2 other files, including reference files (for example `references/check-immutable-rule.md`).

It sits in Backend & APIs, covering Smart contracts. It works with Solidity. The repository describes itself as: Diffyscan detects deployment drift by comparing GitHub source with explorer-verified source and on-chain bytecode across EVM networks. The licence is MIT.

When your agent uses it

  • Accepting an expected diff
  • Narrowing an exception
  • Use debug-diff first if its cause is unknown

Example prompts

  • “/allowed-diffs”

Workflow steps

3 steps, taken from the step headings in SKILL.md.

  1. Establish evidence
  2. Choose a rule
  3. Verify the result

What it can do on your machine

Read from SKILL.md and the folder at commit 9fd6833. 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

    Shell commands in SKILL.md call:

    • uv

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

  • Network

    No URLs in SKILL.md. Its commands use uv, which can reach the network depending on how they are called.

    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

Allowed Diffs loads about 1.2k tokens when it runs, and up to ~1.7k if it reads all its reference files. Until then it costs about 70 tokens; SKILL.md has 592 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~70
When it runs · the whole SKILL.md, loaded when a task matches
~1.2k
With references · SKILL.md plus every file in references/, read only if the agent opens them
~1.7k

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

Auto-check passed

The automated check found no risky patterns in SKILL.md.

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); files beside SKILL.md are not scanned.

SKILL.md

The full file from lidofinance/diffyscan at commit 9fd6833, republished under its MIT licence (© lidofinance). 592 words, ~1,244 tokens.

Download SKILL.mdSave it as .claude/skills/allowed-diffs/SKILL.md (or your agent's skills folder). This skill also uses 1 other file; get the full folder from GitHub.
name
allowed-diffs
description
Adds, reviews or tightens Diffyscan allowed_diffs rules for explained source or bytecode differences, including replacing any wildcards with granular rules. Use when accepting an expected diff or narrowing an exception; use debug-diff first if its cause is unknown.

Encode an explained difference without accepting unrelated drift. Run commands from the repository root.

For review-only requests, inspect existing evidence and report findings or proposed rules; edit configs and run live diagnostics only when the task includes those actions.

1. Establish evidence

Read the target config, bytecode comparison and JSON output. Confirm chain, contract, pinned source and why the difference is intended. If unknown, use debug-diff before changing policy.

Use suggested_rule and actual diff evidence as a starting point. Suggestions describe observations; they do not justify accepting them. To replace a wildcard, remove only that rule for a diagnostic run or use a temporary config copy, then inspect uncovered differences. Preserve other rules and contract scope.

A files suggestion can come from a source file missing on GitHub. Resolve the missing file's provenance, path or dependency before accepting the suggestion; it would otherwise allow the entire source omission.

2. Choose a rule

Inspect evaluate_source_rules, evaluate_bytecode_rules and matchers in diffyscan/utils/allowed_diffs.py when coverage is unclear. Rules are alternatives: one rule must cover the differences; separate entries do not accumulate coverage. Combine necessary facets within one rule.

  • Prefer source line_ranges for exact hunk coordinates. Coordinates are 1-based; count: 0 represents insertion/deletion. These rules do not pin line contents: different edits at the same coordinates are also accepted. files accepts future changes throughout named files.
  • Prefer bytecode immutables with exact on-chain values at compiler-derived offsets when those values explain the difference. byte_ranges constrains offsets and lengths but does not pin values there.
  • Add cbor_metadata: true for explained metadata differences, combined with other necessary facets in the same rule.
  • Use constructor_args or constructor_calldata for an explained alternate simulation, respecting mutual exclusion. Verify the resulting runtime; an override alone does not prove a match.
  • Runtime length and string-literal mismatches currently cannot be accepted by granular byte ranges. Report that limitation without silently widening an exception.

Every rule needs a concrete reason describing deployment evidence and intended scope. any: true excludes other facets and accepts future drift. Bytecode any can also suppress deployment-simulation errors, but not arbitrary compilation or calldata failures. Difficulty reproducing bytecode alone does not justify it. If a wildcard is justified within the requested policy change, document the limitation and align KNOWN_WILDCARDS in tests/test_no_wildcard_regression.py. Remove stale registry entries when removing wildcards; do not weaken the guard.

Show full SKILL.md (210 more words)Show less

The exclusivity of any applies inside one allowed_diffs rule. A constructor override in the separate bytecode_comparison section can coexist with an any rule; that combination is broad policy, not a schema conflict.

Constructor overrides inside rules are evaluated only after the base simulation succeeds and still differs. They cannot recover missing base calldata or a reverting base constructor. Evidence-backed overrides in bytecode_comparison supply the base simulation inputs.

3. Verify the result

Apply validate-config, then:

sh
uv run pytest -q tests/test_allowed_diffs.py tests/test_no_wildcard_regression.py tests/test_configs.py
uv run diffyscan path/to/config.yaml --json -E -G

Confirm the intended contract is allowed or exact, matched reason/facets fit the change, and requested contracts and comparisons remain covered. Check JSON status and errors even with exit code 0. For a new granular rule, evaluate an unrelated hunk, byte offset or immutable value against the matcher and confirm it stays rejected, scoped to the rule's intended guarantees.

In the CLI JSON report, use contracts[].bytecode.facets and .reason; matched_facets and matched_rule are internal result fields, not report keys.

For a runnable immutable-rule check, adapt the matcher example. evaluate_bytecode_rules requires a callable analysis provider; each negative case must supply its changed analysis to that callback as well as to the base argument.

Report evidence, accepted scope, remaining drift protection and checks actually run. If live verification is unavailable, state that the rule has not been verified on chain.

© lidofinance, 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 1 other file (references) in .claude/skills/allowed-diffs of lidofinance/diffyscan.

  • SKILL.md
  • references/check-immutable-rule.md

Open the folder on GitHubat commit 9fd6833

Compare with similar skills

Allowed Diffs next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Allowed Diffs compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Allowed Diffs this skilllidofinance/diffyscan142—~1.2kAutomated safety check: PassMIT
Fizz Convertpashov/skills1.2k2 repos~3.7kAutomated safety check: PassMIT
Feynman Auditor0xiehnnkta/nemesis-auditor2441 repos~11kAutomated safety check: PassMIT
Smart Contract Auditgreatpie/smart-contract-audit-skill101—~1.1kAutomated safety check: PassNone
RadarAuditware/radar154—~2.1kAutomated safety check: PassGPL-3.0
Solidity AuditorGabson0x/bountyforge443—~3.7kAutomated safety check: PassNone

Similar skills

  • Fizz Convert

    pashov/skills

    Convert English-language properties in PROPERTIES.md (produced by the Fizz skill) into Solidity assertions inside the existing fuzz harness, then flip their checkboxes.

    1.2k GitHub starsUsed in 2 repos~3.7k tokens
    Backend & APIsAuto-check passed
  • Feynman Auditor

    0xiehnnkta/nemesis-auditor

    Deep business logic bug finder using the Feynman technique. An agent skill from 0xiehnnkta/nemesis-auditor.

    244 GitHub starsUsed in 1 repo~11k tokens
    Backend & APIsAuto-check passed
  • Smart Contract Audit

    greatpie/smart-contract-audit-skill

    Script-backed, out-of-box auditing workflow for Solidity/EVM repositories based on EVMbench detect/patch/exploit methodology.

    101 GitHub stars~1.1k tokensUpdated 7 mo ago
    Backend & APIsAuto-check passed
  • Radar

    Auditware/radar

    Use radar for smart contract security analysis, AST generation, and detection template development.

    154 GitHub stars~2.1k tokensUpdated 1 mo ago
    Backend & APIsAuto-check passed
  • Solidity Auditor

    Gabson0x/bountyforge

    Security audit of Solidity code while you develop. An agent skill from Gabson0x/bountyforge.

    443 GitHub stars~3.7k tokensUpdated 21 days ago
    Backend & APIsAuto-check passed
  • Scv Scan

    kadenzipfel/scv-scan

    Systematically audit Solidity smart contract codebases for security vulnerabilities using a 4-phase approach - load a vulnerability cheatsheet, sweep code with grep and semantic analysis…

    109 GitHub stars~1.4k tokensUpdated 7 mo ago
    Backend & APIsAuto-check passed

More from lidofinance/diffyscan

  • Add Explorer

    lidofinance/diffyscan

    Adds or repairs Diffyscan explorer API routing and response adapters for a new host, chain or payload format.

    142 GitHub stars~1k tokensUpdated yesterday
    Auto-check passed
  • New Config

    lidofinance/diffyscan

    Creates or extends a Diffyscan verification config for a deployed contract or deployment.

    142 GitHub stars~1k tokensUpdated yesterday
    Auto-check passed
  • Validate Config

    lidofinance/diffyscan

    Reviews an existing Diffyscan YAML or JSON config for load errors, runtime prerequisites, pinned sources, address mappings and broad exceptions.

    142 GitHub stars~1.1k tokensUpdated yesterday
    Auto-check passed
  • Debug Diff

    lidofinance/diffyscan

    Diagnoses failed or incomplete Diffyscan runs, unexpected source or bytecode differences, compilation failures and explorer or RPC errors.

    142 GitHub stars~1.2k tokensUpdated yesterday
    Auto-check passed

Works with

Categories

Questions about Allowed Diffs

What does Allowed Diffs do?

Adds, reviews or tightens Diffyscan alloweddiffs rules for explained source or bytecode differences, including replacing any wildcards with granular rules. Allowed Diffs is an agent skill from lidofinance/diffyscan. Adds, reviews or tightens Diffyscan alloweddiffs rules for explained source or bytecode differences, including replacing any wildcards with granular rules.

When should I use Allowed Diffs?

Allowed Diffs fits situations like: accepting an expected diff; narrowing an exception; use debug-diff first if its cause is unknown.

How do I install Allowed Diffs in Claude Code?

Run `npx skills add lidofinance/diffyscan --skill allowed-diffs -a claude-code`. Or copy the skill folder (.claude/skills/allowed-diffs in lidofinance/diffyscan) into .claude/skills/allowed-diffs in your project. Claude Code loads it when a task matches its description.

How do I install Allowed Diffs in Codex?

Run `npx skills add lidofinance/diffyscan --skill allowed-diffs -a codex`. Or copy the skill folder (.claude/skills/allowed-diffs in lidofinance/diffyscan) into .agents/skills/allowed-diffs in your project. Codex loads it when a task matches its description.

Can I use Allowed Diffs in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add lidofinance/diffyscan --skill allowed-diffs -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/allowed-diffs, .gemini/skills/allowed-diffs, .github/skills/allowed-diffs and .opencode/skills/allowed-diffs in your project.

What does Allowed Diffs need to run?

Going by SKILL.md and its folder, Allowed Diffs needs the command-line tools its instructions call (uv).

Does Allowed Diffs access the network?

SKILL.md contains no URLs. Its commands use uv, which can reach the network depending on how they are called. This is read from the text; nothing was executed.

Is Allowed Diffs safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Allowed Diffs use?

Allowed Diffs is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Allowed Diffs use?

About 1.2k tokens (SKILL.md is roughly 5k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 414 tokens, read only when the agent opens those files.

What are the alternatives to Allowed Diffs?

Skills that share tags, products or a category with Allowed Diffs: Fizz Convert (pashov/skills, 1.2k stars), Feynman Auditor (0xiehnnkta/nemesis-auditor, 244 stars), Smart Contract Audit (greatpie/smart-contract-audit-skill, 101 stars) and Radar (Auditware/radar, 154 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Allowed Diffs?

lidofinance (a GitHub organization) maintains it in lidofinance/diffyscan, which has 142 GitHub stars. The repository holds 5 skills in this directory. The repository was last updated on October 6, 2026.

Source: lidofinance/diffyscan on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.