From dash-platform
Clone Dash Platform repos and regenerate lexicon/ keyword tables. Use when lexicon needs refreshing or after upstream changes.
How this skill is triggered — by the user, by Claude, or both
Slash command
/dash-platform:update-lexiconThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Regenerate `lexicon/*.md` keyword lookup tables from local clones of Dash Platform repos.
Regenerate lexicon/*.md keyword lookup tables from local clones of Dash Platform repos.
bash scripts/clone-repos.sh
Clones/updates all source repos into .repos/ (gitignored). Note commit SHAs from output for lexicon headers.
All content at dashpay.github.io/platform is generated from dashpay/platform — read source directly, no WebFetch needed.
lexicon/rust.mdRust lexicon is generated deterministically from rustdoc JSON — no agent needed.
bash scripts/gen-rustdoc-json.sh
Note the DOC_DIR=... line from output, then:
python3 scripts/gen-rust-lexicon.py "$DOC_DIR" --patterns lexicon/rust-patterns.md > lexicon/rust.md
Requires cargo +nightly (only for this step).
Spawn 4 agents in parallel (subagent_type: "general-purpose"). Each reads source files locally from .repos/.
Critical: Be exhaustive. Read every source file in assigned packages. Extract every public type, function, trait, struct, enum, const, service, RPC, example. Aim for 50+ entries per section. Coverage > speed.
lexicon/rust-patterns.mdPaths: .repos/platform/packages/rs-sdk/src/platform/ (Fetch, FetchMany, Put traits, queries, transitions, builders, mock), .repos/platform/packages/rs-sdk/examples/, .repos/platform/packages/rs-sdk/tests/, .repos/dash-evo-tool/src/ (real-world usage).
Extract: SDK usage patterns (trait conventions, builder patterns, error handling, async, proof verification, mock testing) and examples (every example file, every significant dash-evo-tool backend task). Output only ## Patterns and ## Examples tables — no Types/Functions (those come from the script). Aim for 15+ patterns and 20+ examples.
After this agent completes, re-run the gen-rust-lexicon.py script to concatenate:
python3 scripts/gen-rust-lexicon.py "$DOC_DIR" --patterns lexicon/rust-patterns.md > lexicon/rust.md
lexicon/contract.mdPaths: .repos/platform/packages/rs-dpp/ (contract types, document types, validation), .repos/platform/packages/wasm-dpp/ (WASM bindings), .repos/yappr/contracts/ (contract examples), .repos/dash-bridge/ (TS types).
Extract: Every public type related to data contracts, document types, state transitions, validation, JSON Schema, token config, group actions. Every contract JSON example.
lexicon/js.mdPaths: .repos/platform/packages/js-evo-sdk/, .repos/platform/packages/wasm-dpp/, .repos/platform/packages/wasm-dpp2/, .repos/platform/packages/wasm-sdk/, .repos/yappr/ (TS files), .repos/dash-bridge/ (TS files), .repos/evo-sdk-website/.
Extract: Every exported class, function, type, interface, facade method. Usage patterns from example apps. WASM init patterns.
lexicon/grpc.mdPaths: .repos/platform/packages/dapi-grpc/protos/ (all .proto files).
Extract: Every service, RPC, message type, enum, field. Single package — be completely exhaustive. Every RPC must have a row.
After lexicon tables are regenerated, run the /update-api-changelog skill. This step is not optional — the lexicon and the changelog are released together.
Invoke /update-api-changelog and follow its procedure (plan → gather → synthesis → prune); the full runnable commands and the synthesis-agent prompt live in skills/update-api-changelog/SKILL.md. Existing changelog files are skipped (idempotent), so steady-state runs are cheap.
lexicon/changelog/platform/*.md is produced by this step and committed alongside the rest of lexicon.
Include in each agent's prompt:
Working directory: {plugin_root}
Local clones are in
.repos/. Read source files directly — nogh apior WebFetch. Walk every file in assigned paths using Glob and Read.Be exhaustive. 50+ rows per section minimum. Read every file, not just top-level modules.
Write result to
{output_file}using the format and link prefixes below.
# <Stack> Lexicon
<!-- Auto-generated by update-lexicon. Do not edit. Updated: YYYY-MM-DD -->
<!-- Sources: repo@commit, ... -->
## Types
| Keyword | Description | Src | Docs |
|---------|-------------|-----|------|
## Functions
| Keyword | Description | Src | Docs |
|---------|-------------|-----|------|
## Patterns
| Keyword | Description | Example |
|---------|-------------|---------|
## Examples
| Keyword | Description | File |
|---------|-------------|------|
| Pre | Expands to |
|---|---|
P: | https://github.com/dashpay/platform/blob/master/packages/ |
R: | https://dashpay.github.io/platform/api/rust/ |
G: | https://dashpay.github.io/platform/api/grpc/ |
B: | https://dashpay.github.io/platform/ |
T: | https://github.com/dashpay/dash-evo-tool/blob/master/ |
Y: | https://github.com/PastaPastaPasta/yappr/blob/master/ |
D: | https://github.com/PastaPastaPasta/dash-bridge/blob/master/ |
E: | https://github.com/dashpay/evo-sdk-website/blob/master/ |
npx claudepluginhub lklimek/agents --plugin claudashProvides behavioral guidelines to reduce common LLM coding mistakes, focusing on simplicity, surgical changes, assumption surfacing, and verifiable success criteria.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.