Claude Plugins Community Marketplace

repository·main·Indexed 43 days ago

https://github.com/anthropics/claude-plugins-community

A curated, security-scanned mirror of community-contributed plugins for Claude Cowork and Claude Code. Includes tools such as the eli5 plugin for simplified explanations and the QuickDesign CLI for AI-driven image, video, and marketing asset generation.

Tokens
84.8K
Snippets
152
Records
346
Agent score
92%

What's inside Claude Plugins Community

  1. Overview of TRES Invoice/Bill Matching & ERP Sync

    main

    The tres-invoice-bill-matching skill provides an end-to-end workflow for matching TRES ledger transactions to ERP invoices or bills (Accounts Payable/Accounts Receivable). Once matched, the user can optionally sync the matched entry to a connected ERP such as Xero, QuickBooks Online (QBO), or NetSuite.

    Triggering the skill: Trigger this skill when a user expresses intent to match, link, close, reconcile, or sync an invoice/bill against a blockchain transaction.

    Example trigger phrases:

    • "match this invoice to a transaction"
    • "close invoice INV-123"
    • "link this tx to a bill"
    • "pay this invoice from this transaction"
    • "find the invoice/bill for this hash"

    When NOT to trigger:

    • For general transaction explanations (use tres-tx-story).
    • For ingesting explorer links into the ledger (use tres-explorer-tx-to-ledger).
    • For ERP connection setup (use tres-settings-management).
  2. Overview of the QuickDesign CLI skill

    main
    The quickdesign skill enables Claude to plan and execute AI media generation using the quickdesign CLI. This CLI provides a wrapper around hosted models including Seedance 2.0 (R2V/I2V), Kling, Sora 2, Nano Banana, GPT Image, and video upscale services. It allows for the generation of videos and images directly from a Bash environment without requiring the user to manually manage API keys, polling, or storage uploads.
  3. Overview of the tres-erp-rule-suggestions skill

    main

    The tres-erp-rule-suggestions skill is a user-invocable guide designed to help crypto organizations on the Tres Finance platform create, review, or fix ERP accounting rules. It maps transaction types to the correct chart of accounts entries using a layered rule engine.

    Trigger: The skill is activated when a user asks to create, review, or fix ERP rules or accounting mappings for an organization.

    Data Access: All data fetching, validation, and rule application are performed via the tres-finance MCP server. No direct database access or external scripts are required.

    name: tres-erp-rule-suggestions
    description: >
      Guide users through creating ERP accounting rules for crypto organizations on the Tres Finance platform.
      Maps every transaction type to the correct chart of accounts entries using a layered rule engine.
      Trigger: user asks to create, review, or fix ERP rules / accounting mappings for an organization.
    user_invocable: true
  4. Capabilities of TestDino Plugins

    main

    The TestDino MCP server allows AI agents to interact with your test data and automation suites. Key capabilities include:

    • Connection & Access: Validate access and discover organizations/projects.
    • Test Run Analysis: List runs and review failures filtered by branch, commit, author, environment, or time range.
    • Test Case Debugging: Inspect failure details, historical failures, retries, and artifacts.
    • Manual Testing: Create/update manual test cases, suites, and runs (including results).
    • Releases: Manage releases and milestones.
    • Exploratory Sessions: Create and track exploratory testing sessions.
    • Audits: Run the TestDino audit flow on Playwright test code.
  5. TestDino session tool capabilities

    main

    The testdino-sessions skill provides the following tool capabilities for managing exploratory testing sessions:

    ToolPurpose
    list_sessionsBrowse sessions by state, status, session type, assignee, release, or tags.
    get_sessionInspect a specific session in detail.
    create_sessionCreate a new exploratory testing session.
    update_sessionUpdate the properties of an existing session.
    healthUsed to verify connectivity and retrieve projectId if unknown.
  6. Understand the TRES Finance Rollup Rules skill

    main

    The tres-rollup-rules skill is a data-driven discovery tool used to analyze sub-transaction patterns in TRES Finance. It identifies repetitive transaction groups and proposes 'rollup rules' to consolidate them into single aggregated entries (daily or monthly) for cleaner ERP integration.

    Key Concepts:

    • Rollup Rule Definition: A rule is uniquely identified by four dimensions: internalAccount (wallet) + asset KEY + platform + balanceFactor (direction). You can also add optional discriminating filters like fees, methodIds, subtxType, amount bounds, or counterparties.
    • Non-destructive: Rollups do not delete data; original raw sub-transactions remain visible in a Rollup Breakdown report.
    • Discovery Flow: The skill follows a pipeline of confirming scope, inventorying existing data/rules, and performing core analysis to find candidates.
    • Compatibility: Requires the TRES Finance MCP connector.
  7. Default Video Model: Use seedance-2.0-r2v

    main

    For any UGC (User Generated Content), talking-avatar, promo, explainer, or multi-scene work, always use the seedance-2.0-r2v model.

    Do not downgrade to seedance-2.0-i2v even if the script is short. The R2V model is required to support the core primitives of this skill, including:

    • @Image1 references
    • Multi---reference-image support
    • --reference-audio for voice continuity

    Only switch away from R2V if the user explicitly requests a different model (e.g., "use Sora 2") or if R2V is unavailable in the registry.

  8. Data formats and identifiers for TestDino manual runs

    main

    When calling TestDino manual-run tools, ensure you use the correct formats for identifiers and payloads:

    Identifiers

    • runId: Accepts an internal ID or a counter-style ID (e.g., RUN-12).
    • rtcRef: Accepts a run test case row ID, a case key (e.g., TC-156), or the underlying test case ID.
    • updates.assigneeUserId: Accepts a user ID or an email address.

    Payload Requirements

    • updates.result: While display forms are accepted, use these canonical values for safety: passed, failed, blocked, skipped, retest, or untested.
    • Arrays: For both create and update payloads, the tags and linkedIssues fields must be provided as JSON arrays.
  9. Construct sub-transactions for the ledger

    main

    When building the subTransactions array for the createManualTransactionWithSubTransactions mutation, follow these direction and action rules:

    ScenarioDirectionFinancial Action
    User's wallet sends tokensOUTFLOWTOKEN_TRANSFER
    User's wallet receives tokensINFLOWTOKEN_TRANSFER
    Gas fee (paid by sender)OUTFLOWGAS
    User's wallet sends native currencyOUTFLOWNATIVE_TRANSFER
    User's wallet receives native currencyINFLOWNATIVE_TRANSFER

    Key Rules:

    • Token transfers: One sub-transaction per event. thirdPartyIdentifier is the counterparty address.
    • Gas fee: Always create a GAS sub-transaction if the user is the sender. Set thirdPartyIdentifier to the fixed value "Native".
    • Internal transfers: If the user owns both the sender and receiver addresses, create two sub-transactions (one OUTFLOW, one INFLOW) with their respective belongsToIds, plus a GAS sub-transaction for the sender.
    • Fiat values: Do not pass fiatValue unless explicitly requested; the system handles pricing automatically.
  10. Avoid avatar regeneration using edit-style prompts

    main

    A common failure mode is "Compose-mode," where the model regenerates a fresh AI avatar instead of editing the provided source. This happens if the prompt starts with descriptive composition phrases (e.g., "Compose a vertical 9:16 UGC selfie frame..."). This results in a loss of the source's lighting, grain, and authenticity.

    To ensure an edit rather than a regeneration:

    • Use edit-style prompts: Edit @Image1: add ... or Take @Image1 as-is and only change ....
    • Do not re-list scene tokens already present in the reference image.
    • Strip quality-upgrade phrases like photo-realistic, studio quality, or 8K, as these trigger regeneration.
  11. How TRES Settings Management works

    main

    The tres-settings-management skill allows for viewing and modifying configurations at two distinct levels via the TRES MCP GraphQL API:

    1. Organization Settings: Global configurations that apply to the entire organization (e.g., cost basis calculation methods, ERP behavior, pricing sources, and feature flags).
    2. Platform Settings: Granular configurations that apply to specific blockchain or exchange platforms (e.g., Ethereum, Arbitrum) and can optionally be scoped to specific wallets (e.g., commit strategies, sync date boundaries, and asset filtering).

    When interacting with this skill, the system follows a strict 'Read-Before-Write' workflow and requires explicit user approval for any mutations.

  12. Understand the exported XLSX workbook structure

    main

    The exported workbook contains two specific tabs designed for different stages of the contact building process:

    1. Contacts (Import-Ready)

    This sheet matches the TRES contacts import template exactly. It is intended for the user to fill out and then upload back into TRES using the contacts import skill.

    ColumnDescription
    Contact NameLeave blank; user fills this in.
    Contact AddressThe address identifier (original casing).
    Contact TagLeave blank; user fills this in (e.g., Exchange, Vendor).

    2. Address details (Enrichment)

    This is a reference tab used to help users prioritize which addresses to label. It contains activity statistics and inferred network information.

    ColumnDescription
    Contact AddressThe address identifier.
    NetworkThe inferred network (e.g., EVM, Tezos, Tron).
    Inflow TxsCount of incoming transactions.
    Outflow TxsCount of outgoing transactions.
    Inflow USDTotal USD value of inflows.
    Outflow USDTotal USD value of outflows.
    Total USDSum of inflow and outflow USD values.

    Sorting: Both tabs are sorted by Total USD (inflow + outflow) in descending order so that the most active counterparties appear first.