SKILLEMALL.ai
GitHub ↗ RU Sign in

CD torch-liquidation-bot

Autonomous vault-based liquidation keeper for Torch Market lending on Solana. Scans all migrated tokens for underwater loan positions (LTV > 65%) using the SDK's built-in bulk loan scanner (getAllLoanPositions), builds and executes liquidation transactions through a Torch Vault, and collects a 10% collateral bonus. The agent keypair is generated in-process -- disposable, holds nothing of value. All SOL and collateral tokens route through the vault. The human principal creates the vault, funds it, links the agent, and retains full control. Built on torchsdk v3.7.22 and the Torch Market protocol.

modbender/skill-library-mcp Claude Code author: modbender MIT 21 files body ≈ 4 112 tokens Open the sourcegithub.com↗ analyzed 2 d ago

Autonomous vault-based liquidation keeper for Torch Market lending on Solana.

As a process D 47/100 · Unfinished process — weak spots: result and completion, inputs and preconditions, consistency

ProcedureGitHubAI and agentstype and topics are labelled automatically from the skill text
Runs in: OpenClaw Claude / Claude Code ChatGPT / Codex Gemini CLI Cursor GitHub Copilot / VS Code Qwen Code Kimi Code CLI Z.ai (GLM) DeepSeek (Deep Code) Mistral Vibe LangChain (deepagents) Goose OpenCode
JSON
Technical rating
C
73/100
safety, quality, tests
Safety 60%
74
Quality 40%
72
Run on models
none yet
Process rating
D
47/100
Unfinished process
Result and completion w 14
0
Inputs and preconditions w 11
0
Running it twice w 4
30
the three weakest of ten parameters · all ten

What is at stake

Medium-severity findings: the skill is probably honest, but read what alarmed the scanner.

Risky intent medium severity

Below is the worst case for this category. The finding here is medium: the guard saw a sign, not a proof.

If you install

The purpose itself is risky: wallets, browser password stores, offensive security. Even an honest implementation gives the agent access to things that cost money.

For the author

Explain in the description why the access is needed and how it is limited; add tests that show refusals on dangerous requests.

How to improve

  1. Say in the description WHEN to use the skill ("use when…", example requests): that is the agent's main cue.
For the model run — optional
  • Your own cases (evals/evals.json, 4–6 real requests with expected answers): the full check would then run those instead of a model-drafted suite.
  • A spec.yaml with trigger phrases and assertions — a behaviour contract for CI; `skilltest init` writes a template.

Guard findings · 22

✓ No critical or high findings

Medium and low: 22
  • medium Risky intent intent-wallet-secrets SKILL.md:369
    Handles crypto-wallet secrets (seed / mnemonic / private key) — a classic stealer target
    1. **Never ask a user for their private key or seed phrase.** The vault authority signs from their own device.
  • low Secrets in code secret-high-entropy-token agent.json:61
    High-entropy token-like string (may be an id, hash or a credential) (quoted — discussed, not commanded)
    "program_id": "8hbU…BeT",
    quoted
  • low Secrets in code secret-high-entropy-token agent.json:165
    High-entropy token-like string (may be an id, hash or a credential) (quoted — discussed, not commanded)
    "wallet": "8cpW…nMa",
    quoted
  • low Secrets in code secret-high-entropy-token agent.json:166
    High-entropy token-like string (may be an id, hash or a credential) (detector / deny-list definition)
    "pda": "2rNk…Mmb",
    detector
  • low Secrets in code secret-high-entropy-token audit.md:8
    High-entropy token-like string (may be an id, hash or a credential) (quoted — discussed, not commanded)
    **On-Chain Program:** `8hbU…BeT` (V3.7.7, 27 instructions)
    quoted
  • low Secrets in code secret-high-entropy-token audit.md:469
    High-entropy token-like string (may be an id, hash or a credential) (quoted — discussed, not commanded)
    **On-Chain Version:** V3.7.7 (Program ID: `8hbU…BeT`, 27 instructions)
    quoted
  • low Secrets in code secret-high-entropy-token design.md:9
    High-entropy token-like string (may be an id, hash or a credential) (detector / deny-list definition)
    The bot is built on `to…@….7.22` and targets the Torch Market on-chain program (`8hbU…BeT`). It uses the SDK's bulk loan scanner (`getAllLoanPositions`) to dis
    detector
  • low Secrets in code secret-high-entropy-token lib/torchsdk/constants.js:6
    High-entropy token-like string (may be an id, hash or a credential) (quoted — discussed, not commanded)
    exports.PROGRAM_ID = new web3…Key('8hbU…BeT');
    quoted
  • low Secrets in code secret-high-entropy-token lib/torchsdk/constants.js:14
    High-entropy token-like string (may be an id, hash or a credential) (quoted — discussed, not commanded)
    ? 'CPMD…kCW'
    quoted
  • low Secrets in code secret-high-entropy-token lib/torchsdk/constants.js:15
    High-entropy token-like string (may be an id, hash or a credential) (quoted — discussed, not commanded)
    : 'CPMM…P1C');
    quoted
  • low Secrets in code secret-high-entropy-token lib/torchsdk/constants.js:20
    High-entropy token-like string (may be an id, hash or a credential) (quoted — discussed, not commanded)
    exports.WSOL_MINT = new web3…Key('So11…112');
    quoted
  • low Secrets in code secret-high-entropy-token lib/torchsdk/constants.js:23
    High-entropy token-like string (may be an id, hash or a credential) (quoted — discussed, not commanded)
    ? '9zSz…Bj6'
    quoted
  • low Secrets in code secret-high-entropy-token lib/torchsdk/tokens.js:412
    High-entropy token-like string (may be an id, hash or a credential) (detector / deny-list definition)
    programId === 'Memo…cHr' ||
    detector
  • low Secrets in code secret-high-entropy-token lib/torchsdk/torch_market.json:2
    High-entropy token-like string (may be an id, hash or a credential) (quoted — discussed, not commanded)
    "address": "8hbU…BeT",
    quoted
  • low Secrets in code secret-high-entropy-token lib/torchsdk/torch_market.json:914
    High-entropy token-like string (may be an id, hash or a credential) (quoted — discussed, not commanded)
    "address": "ATok…knL"
    quoted
  • low Secrets in code secret-high-entropy-token lib/torchsdk/torch_market.json:1302
    High-entropy token-like string (may be an id, hash or a credential) (quoted — discussed, not commanded)
    "address": "ATok…knL"
    quoted
  • low Secrets in code secret-high-entropy-token lib/torchsdk/torch_market.json:1310
    High-entropy token-like string (may be an id, hash or a credential) (quoted — discussed, not commanded)
    "address": "Sysv…111"
    quoted
  • low Secrets in code secret-high-entropy-token lib/torchsdk/torch_market.json:1803
    High-entropy token-like string (may be an id, hash or a credential) (quoted — discussed, not commanded)
    "address": "ATok…knL"
    quoted
  • low Secrets in code secret-high-entropy-token SKILL.md:45
    High-entropy token-like string (may be an id, hash or a credential)
    program-id: 8hbU…BeT
  • low Secrets in code secret-high-entropy-token SKILL.md:446
    High-entropy token-like string (may be an id, hash or a credential) (quoted — discussed, not commanded)
    - Program ID: `8hbU…BeT`
    quoted
  • low Risky intent intent-offensive-security verification.md:183
    Offensive-security / dual-use content (legitimate for authorised testing; review intended use)
    | **CPI safety** | Reentrancy via Raydium CPIs, privilege escalation | Cross-program invocation is outside arithmetic scope |
  • low Risky intent intent-offensive-security verification.md:194
    Offensive-security / dual-use content (legitimate for authorised testing; review intended use)
    3. **CPI safety** -- can Raydium CPIs be exploited for reentrancy or privilege escalation?

Files scanned: 21. Evidence is masked. Grey chips explain why severity was lowered.

Against the Agent Skills spec

  • warning description-no-when description does not say WHEN to use the skill (no "use when")
  • note frontmatter-key unknown frontmatter key "requires"

Process rating: all ten parameters 47/100

  • 0Result and completion. Does not say what the result is
  • 0Inputs and preconditions. Does not say what the process needs to start
  • 30Running it twice. 2 mutating operations with no state check
  • 40Consistency. Frontmatter name (torch-liquidation-bot) differs from the folder (torchliquidationbot)
  • 50When it triggers. No condition that starts the skill
  • 50Failures and branches. 0 branches, has a failure section
  • 60Tools and files. Uses tools (web) that frontmatter does not declare
  • 70Execution cost. Instruction body is 4112 tokens
  • 100Steps. 30 steps
  • 100Progress reporting. Reports progress
  • medium Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
  • low 11 top-level sections: this looks like several domains in one skill

Everything here is measured from the skill text rather than judged by a model, so the numbers are checkable. A parameter weighs more when it is a more common reason for the process to stall.

Quality signals

  • +5Description has no quoted example phrases that should trigger the skill
  • +4Description does not say when NOT to use the skill (false activations)
  • +3Output format is not stated: the model decides each time
  • +2Single-language instructions
  • +3Description length 601: enough signal without eating the budget
  • +4Structure: 26 headings
  • +3Step-by-step instructions: 30 items
  • +4Has examples (9 code blocks)
  • +1License stated

Quality base 70; lint remarks subtract, signals add up to 100. Result: 72.