claude-extensions

Comprehensive security scan for any Solidity project (Foundry, Hardhat, or bare). Orchestrates parallel agents — defensive analysis, adversarial exploit hypotheses, gas optimization, and optional deployment-script auditing — into one consolidated report. Trigger on: "solidity security scan", "audit solidity", "scan contracts", "full solidity review".

Trace code flow as an indented-outline document where indentation marks logic forks. Produces a flow doc with an overview, a coupling/wiring table (channels, shared state, events), per-flow outlines citing file:line on every node, and a goal-driven analysis section (condensation opportunities by default). Use this whenever the user wants to understand how a module/crate/subsystem flows, says code "feels loosely organized" or "hard to follow", wants to "track the flow", "outline the logic", "map the call paths", "see every branch/fork", or wants to understand code well enough to condense, simplify, or refactor it — even if they don't say the word "outline". Also trigger when the user asks for a document tracing which events cause which functions to run.

Interactive consultant that interviews the user, recommends configuration, and generates a Claude Code agent definition with hard responsibility boundaries, domain context, quality gates, and optional persistent memory. Trigger on: "create agent", "new agent", "design an agent", "build an agent", "I need an agent that", "agent for", "set up an agent". Do NOT trigger for skill creation — use skill-creator instead.

Deep combined audit using iterative Feynman + State Inconsistency analysis across 8 phases with specialized agents. Language-agnostic. Dynamically discovers domain-specific patterns via Phase -1 before the main pipeline. Spawns nemesis-orchestrator which coordinates domain discovery, recon, mapping, interrogation, state checking, feedback loop, journey tracing, verification, and reporting. Triggers on /nemesis-scan or deep combined audit.

Generate Criterion benchmarks for telcoin-network hot paths. Trigger on: "benchmark", "add bench", "performance test", "criterion", "measure latency", "throughput test"

Deep bug hunt for the telcoin-network codebase (Rust + Solidity submodule) across 9 phases with specialized agents. Bug-hunter framing — asks 'how does this fail in production?' not 'how does an attacker exploit this?'. Covers concurrency, determinism, consensus correctness, state atomicity, panic surface, fork risk, and error propagation. Produces bug tickets with repro + failure mode per finding. Triggers on /tn-bug-scan or tn bug scan or telcoin bug hunt.

Perform automated security hardening sweeps on the telcoin-network codebase. Trigger on: "harden", "security check", "determinism check", "find unwrap", "production ready", "audit prep", "hardening sweep", "panic audit", "tracing audit", "blocking audit", "review"

Deep combined audit using iterative Feynman + State Inconsistency analysis across 8 phases with specialized agents. Language-agnostic. Dynamically discovers domain-specific patterns via Phase -1 before the main pipeline. Spawns nemesis-orchestrator which coordinates domain discovery, recon, mapping, interrogation, state checking, feedback loop, journey tracing, verification, and reporting. Triggers on /nemesis-scan or deep combined audit.

Code Review & Security Analysis for tn-contracts

Code Review & Security Analysis Skill

Reference skill containing all telcoin-network Rust coding conventions, rules, and anti-patterns. Invoked by tn-* subagents to load domain knowledge into their context window. NOT user-invocable. Invoked programmatically by tn-* agents via the Skill tool.

Comprehensive security evaluation for telcoin-network PRs and branches. Orchestrates 10 parallel agents covering consensus safety, state transitions, crypto, DoS, determinism, contract safety, dependency audit, nemesis logic audit, DREAD scoring, and STRIDE classification, with independent verification to eliminate false positives. Trigger on: "security eval", "security audit PR", "is this PR safe", "pre-merge security".

Generate security threat model documentation for telcoin-network components. Trigger on: "threat model", "attack surface", "security architecture", "adversary model", "security doc", "audit prep"

Generate crate-level documentation for telcoin-network crates. Trigger on: "document this crate", "write docs", "add rustdoc", "architecture doc", "module docs", "crate docs", "write crate documentation"

Design and generate end-to-end tests for the telcoin-network node. Trigger on: "write e2e test", "integration test", "test epoch", "test restart", "test sync", "end to end"

Generate property-based tests using proptest for the telcoin-network codebase. Trigger on: "write proptest", "property test", "invariant test", "fuzz this", "test properties"

Generate two markdown files for the current branch — a focused GitHub issue (Problem/Solution) and a broad PR comment summarizing all changes. Trigger on "create an issue", "gh issue", "draft issue", "issue for this branch", or when the user wants to describe their branch changes for reviewers. Do NOT trigger for posting to GitHub via the API — this skill writes local markdown files only.

Style guide for writing prose that sounds human, not AI-generated. Use whenever Claude writes or edits prose — markdown, GitHub issues, PR descriptions, commit messages, code comments, documentation, changelogs, explanations. Also trigger on "write naturally", "sound human", "avoid AI tone". Applies when other writing skills like gh-issue produce output. Do NOT trigger for pure code generation or structured data.

Debug failing end-to-end tests in the telcoin-network repo (Narwhal/Bullshark consensus + Reth EVM). Trigger when the user shares e2e stdout/stderr, mentions a failing e2e test, asks about test_logs, or pastes node traces, consensus errors, or execution-engine failures. Covers panics, timeouts, races, epoch boundaries, restart failures, consensus hangs.

Domain expert reference for the telcoin-network BFT consensus layer — Bullshark ordering, certificate construction and validation, vote aggregation, header chains, DAG invariants, and quorum math. Loaded by tn-rust-engineer and tn-domain-reviewer when work touches the primary, certifier, proposer, executor (consensus output), or aggregator code paths. NOT user-invocable. Loaded programmatically by tn-* agents via the Skill tool.

Domain expert reference for the telcoin-network smart-contract integration layer — ConsensusRegistry, StakeManager, Issuance bindings; system calls (concludeEpoch, applyIncentives, applySlashes); validator activation/exit; reward tier accounting. Loaded by tn-rust-engineer and tn-domain-reviewer when work touches crates/tn-reth/src/system_calls.rs, system-call ordering, or registry reads from Rust. NOT user-invocable. Loaded programmatically by tn-* agents via the Skill tool.

Domain expert reference for the telcoin-network epoch lifecycle and node-management layer. Loaded by tn-rust-engineer and tn-domain-reviewer when work touches epoch boundaries, EpochManager, RunEpochMode, GasAccumulator catchup, governance-driven config updates, or any state read/written at the start or end of an epoch. NOT user-invocable. Loaded programmatically by tn-* agents via the Skill tool.

Domain expert reference for the telcoin-network execution layer — Reth integration, EVM block production, payload building, and the executor/engine boundary contract. Loaded by tn-rust-engineer and tn-domain-reviewer when work touches reth_env, payload builder, batch builder's block-shaping logic, base fee derivation, or any code that reads or writes EVM state. NOT user-invocable. Loaded programmatically by tn-* agents via the Skill tool.

Domain expert reference for the telcoin-network libp2p layer — gossipsub topics, request-response protocols, peer discovery, epoch-aware filtering, ConsensusNetwork lifecycle. Loaded by tn-rust-engineer and tn-domain-reviewer when work touches crates/network-libp2p/**, crates/state-sync/**, peer/topic management, or any code that publishes/subscribes/requests over the wire. NOT user-invocable. Loaded programmatically by tn-* agents via the Skill tool.

Domain expert reference for the telcoin-network storage layer — consensus DB (REDB), reth-db (MDBX), key encoding, table layout, atomic writes, epoch-scoped vs persistent tables. Loaded by tn-rust-engineer and tn-domain-reviewer when work touches the Database trait, table definitions, snapshot logic, or any read/write to the consensus DB. NOT user-invocable. Loaded programmatically by tn-* agents via the Skill tool.

Domain expert reference for the telcoin-network worker layer — batch construction, transaction pool management, EIP-1559 fee calculation, beneficiary committee enforcement, and the worker/primary boundary contract. Loaded by tn-rust-engineer and tn-domain-reviewer when work touches the worker crate, batch-builder, transaction pool, batch fetcher, or quorum-waiter. NOT user-invocable. Loaded programmatically by tn-* agents via the Skill tool.

Orchestrator for Rust development in the telcoin-network repo. Spawns subagents for task analysis, implementation planning, code writing, and verification. Trigger on: "implement", "fix", "add", "build", "refactor", "new feature", "bug fix", "extend", "port", "migrate", "wire up", "integrate". Do NOT trigger for: code review (tn-review), e2e debugging (tn-debug-e2e), contract review (tn-review-contracts), writing tests (tn-write-e2e / tn-write-proptest).

Sequential editing pipeline for technical documentation. Decomposes human-writing guidelines into focused single-pass agents that each apply one narrow set of rules. Works for any professional technical documentation: architecture docs, READMEs, crate docs, guides, changelogs. Trigger on: "doc-writer", "edit docs", "polish docs", "run doc pipeline", "clean up this document", "make this sound human"

The Inescapable Auditor. Runs the full Feynman Auditor (Stage 1) and full State Inconsistency Auditor (Stage 2) as primary steps, then fuses their outputs in a feedback loop (Stage 3) to find bugs at the intersection that neither alone would catch. Language-agnostic. Triggers on /nemesis or nemesis audit.

Deep business logic bug finder using the Feynman technique. Language-agnostic — works on Solidity, Move, Rust, Go, C++, or any codebase. Questions every line, every ordering choice, every guard presence/absence, and every implicit assumption to surface logic bugs that pattern-matching misses. Triggers on /feynman, feynman audit, or deep logic review.

Finds state inconsistency bugs where an operation mutates one piece of coupled state without updating its dependent counterpart, causing silent data corruption or reverts in subsequent operations. Triggers on /state-audit, state inconsistency audit, or coupled state audit.

Guide for creating mermaid diagrams. This skill should be used when users want to create a mermaid diagram (or update an existing diagram).