Not in the top 1,000 by GitHub stars
aaronjmars/agent-credit
Give your AI agent a credit line: it borrows and repays on Aave V3 via credit delegation, within limits you set.
History
Review weekly changes to SKILL.md before updating. Dates mark the weeks in which we observed each version; checks apply to the recorded commits.
View weekly changes
SKILL.md
Week of Nov 2, 2026 · Line counts unavailable
Before Not checked yet → After Network or broad shell
First observation → ac91a2d8f30d
No earlier version was observed for comparison.
Heat
2.7- Star growth, last 7 daysNot available
- Star growth, last 24 hoursNot available
- Rank movement, last 7 daysNot available
- Commits, last 30 daysNot available
- Accounts asking about it through MCP, last 7 daysNot available
- Hot votes this weekNot available
Heat never changes the order by stars, and sponsorship never touches it. How heat works
Stars over time
30-day change not yet available: not enough history.
Individual skills in this repository
How it works
agent-credit
When your agent uses it
Borrow from Aave V3 via credit delegation. Agent draws against delegator collateral. Supports borrow, repay, health checks.
What it loads into context
Frontmatter
32 tokensAlways available to the agent
SKILL.md
3,948 tokensWhen the skill is activated
Reference documents
13,087 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- Compatible With
- How Credit Delegation Works
- Flow
- Quick Start
- Core Usage
- Safety System
- Capabilities
- Supported Chains
- Common Patterns
- Configuration Reference
- Error Handling
- Security
- Resources
Scripts executed separately
aave-borrow.shaave-repay.shaave-setup.shaave-status.shlib.shtests/mocks/casttests/test-borrow-contract.shtests/test-repay-contract.sh
Files
.github/CONTRIBUTING.mddoc4,240 bytes.github/README.mddoc14,039 bytes.github/SECURITY.mddoc5,162 bytes.github/workflows/ci.ymlasset1,887 bytes.gitignoreasset212 bytesLICENSEasset1,067 bytesSKILL.mdcore15,599 bytesSKILL.mdcore15,599 bytesSKILL.mdcore15,599 bytesSKILL.mdcore15,599 bytesSKILL.mdcore15,599 bytesSKILL.mdcore15,599 bytesSKILL.mdcore15,599 bytesSKILL.mdcore15,599 bytesSKILL.mdcore15,599 bytesSKILL.mdcore15,599 bytesSKILL.mdcore15,599 bytesSKILL.mdcore15,599 bytesSKILL.mdcore15,599 bytesSKILL.mdcore15,599 bytesSKILL.mdcore15,599 bytesSKILL.mdcore15,599 bytesSKILL.mdcore15,599 bytesSKILL.mdcore15,599 bytesSKILL.mdcore15,599 bytesaave-borrow.shruns15,407 bytesaave-repay.shruns8,176 bytesaave-setup.shruns7,335 bytesaave-status.shruns6,726 bytesconfig.example.jsonasset749 bytescontracts.mddoc4,479 bytesdeployments.mddoc10,415 bytesdocs/assets/btn-docs.svgasset1,117 bytesdocs/assets/btn-star.svgasset975 bytesdocs/assets/btn-x.svgasset1,143 bytesdocs/assets/hero-animated.svgasset5,516 bytesimg/approve.pngasset111,848 bytesimg/credit.pngasset1,325,939 bytesimg/token-find.pngasset160,241 byteslib.shruns6,122 bytessafety.mddoc5,265 bytestests/mocks/castruns6,900 bytestests/test-borrow-contract.shruns16,633 bytestests/test-repay-contract.shruns9,237 bytes
Read SKILL.md
# Aave Credit Delegation
Borrow funds from Aave using delegated credit. Your main wallet supplies collateral and delegates borrowing power to the agent's wallet. The agent can then autonomously borrow tokens when needed — the debt accrues against the delegator's position.
> **Protocol:** Borrowing requires **Aave V3**. `aave-borrow.sh` resolves the price oracle through `ADDRESSES_PROVIDER()`, which exists on V3's `Pool` but not on V2's `LendingPool` (V2 exposes `getAddressesProvider()`), so a borrow aborts on a V2 market before any safety check runs. `aave-status.sh`, `aave-repay.sh`, and `aave-setup.sh` do not use that getter and do work against V2.
>
> The credit-delegation signatures themselves (`borrow`, `repay`, `approveDelegation`, `borrowAllowance`) are identical across both versions. V2 denominates collateral and debt in ETH (18 decimals) rather than USD (8), so set `AAVE_BASE_CURRENCY_DECIMALS=18` there for correct `~$X` display; the safety comparisons hold either way.
## Compatible With
- **[OpenClaw](https://openclaw.ai/)** — Install as a skill, the agent borrows autonomously
- **[Claude Code](https://www.npmjs.com/package/@anthropic-ai/claude-code)** — Run scripts directly from a Claude Code session
- **Any agent framework** — Plain bash + Foundry's `cast`, works anywhere with a shell
View the complete source on GitHubSkills included
- agent-credit
Repository facts
- License
- MIT
- Created
- Feb 5, 2026
- Last pushed
- Nov 5, 2026
- Forks
- 4
- Open issues
- 0
- Project activity
- Active
Before you install
Network or broad shellAutomated signals, not a guarantee
Checked Nov 6, 2026
Not fully analyzed: some files were too long, unreadable or missing. This never lowers the level.
Show 18 notes
SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedSKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedSKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedSKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedSKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedSKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedSKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedSKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedSKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedSKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedSKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedSKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedSKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedSKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedSKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedSKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedSKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedSKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzed
- No broad tool permissions
- 8 dangerous shell commandsin 5 files: .github/workflows/ci.yml:20 tests/test-borrow-contract.sh:37 tests/test-repay-contract.sh:34 and 2 more
- 70 network referencesin 13 files: .github/README.md:5 SKILL.md:16 .github/CONTRIBUTING.md:21 and 10 more
- No hidden or obfuscated text
- No instruction overrides
- 8 executable scriptsin 8 files: aave-borrow.sh:1 aave-repay.sh:1 aave-setup.sh:1 and 5 more
- 24 references to secretsin 7 files: aave-setup.sh:53 aave-status.sh:128 aave-borrow.sh:141 and 4 more
- No requests to vote in Hot
Show all 110 signals by file
.github/CONTRIBUTING.md
- Network accessline 21
`fix/…`, `docs/…`). - **One change per PR.** Don't bundle unrelated edits. - **Title as a [Conventional Commit](https://www.conventionalcommits.org/)** — - Network accessline 22
- **One change per PR.** Don't bundle unrelated edits. - **Title as a [Conventional Commit](https://www.conventionalcommits.org/)** — `feat: …`, `fix: …`, `docs: …`. PRs are squash-merged, so the title becomes the - Network accessline 23
- **Title as a [Conventional Commit](https://www.conventionalcommits.org/)** — - Network accessline 34
**Prerequisites:** a POSIX shell, [Foundry](https://book.getfoundry.sh/) (`cast`), - Network accessline 38
git clone https://github.com/aaronjmars/agent-credit.git && cd agent-credit
.github/README.md
- Network accessline 5
<p align="center"> <strong>Star us ❤<U+FE0F> →</strong> <a href="https://github.com/aaronjmars/agent-credit/stargazers"><img src="../docs/assets/btn-star.svg" alt="Star Agent Credit on GitHub" height="34" align="absmiddle - Network accessline 6
<strong>Star us ❤<U+FE0F> →</strong> <a href="https://github.com/aaronjmars/agent-credit/stargazers"><img src="../docs/assets/btn-star.svg" alt="Star Agent Credit on GitHub" height="34" align="absmiddle"></a> - Network accessline 7
<a href="https://github.com/aaronjmars/agent-credit/stargazers"><img src="../docs/assets/btn-star.svg" alt="Star Agent Credit on GitHub" height="34" align="absmiddle"></a> - Network accessline 8
<a href="../SKILL.md"><img src="../docs/assets/btn-docs.svg" alt="Agent Credit skill docs" height="34" align="absmiddle"></a> <a href="https://x.com/aaronjmars"><img src="../docs/assets/btn-x.svg" alt="@aaronjmars on X" height - Network accessline 9
<a href="https://x.com/aaronjmars"><img src="../docs/assets/btn-x.svg" alt="@aaronjmars on X" height="34" align="absmiddle"></a> - Network accessline 24
[](https://github.com/aaronjmars/agent-credit/stargazers) - Network accessline 25
[](https://github.com/aaronjmars/agent-credit/network/members) - Network accessline 26
[](https://aave.com) - Network accessline 27
[](../LICENSE) - Network accessline 33
- **[OpenClaw](https://openclaw.ai/)** - Install as a skill and the agent can borrow autonomously - Network accessline 34
- **[Claude Code](https://www.npmjs.com/package/@anthropic-ai/claude-code)** - Run the scripts directly from a Claude Code session - Network accessline 37
Combines naturally with **[Bankr](https://bankr.bot/)** skills - borrow USDC via delegation, then use Bankr to swap, bridge, or deploy it. The agent gets a credit line *and* a full DeFi toolkit. - Dangerous shell operationline 91
curl -L https://foundry.paradigm.xyz | bash && foundryup - Network accessline 91
curl -L https://foundry.paradigm.xyz | bash && foundryup - Access to secretsline 102
[Credential-related source omitted] - Network accessline 133
Go to [app.aave.com](https://app.aave.com), connect your wallet, and supply collateral (ETH, USDC, etc.). This is standard Aave usage - nothing specific to credit delegation yet. - Network accessline 143
On [app.aave.com](https://app.aave.com), go to the reserve page for the asset. Click the token icon to expand a dropdown showing the underlying token, the aToken, and the **debt token** address: - Network accessline 215
On [app.aave.com](https://app.aave.com), your dashboard shows all outstanding borrows. Any debt the agent created shows up here - it's on your position. - Network accessline 219
On [app.aave.com](https://app.aave.com), click **Repay** on any borrow. Standard Aave repayment - the agent doesn't need to be involved. The agent can also repay via `aave-repay.sh`. - Network accessline 225
- **Start small.** Approve $50-100 initially. Increase after you've tested the flow. - **Never approve `type(uint256).max`.** Always set a concrete ceiling per asset. - **Monitor your health factor.** Set up alerts on [app.aave.com](https:/ - Network accessline 226
- **Never approve `type(uint256).max`.** Always set a concrete ceiling per asset. - **Monitor your health factor.** Set up alerts on [app.aave.com](https://app.aave.com) or [DeFi Saver](https://defisaver.com). - **Revoke when idle.** If the - Network accessline 227
- **Monitor your health factor.** Set up alerts on [app.aave.com](https://app.aave.com) or [DeFi Saver](https://defisaver.com). - Network accessline 238
| Action | How | |--------|-----| | Supply collateral | [app.aave.com](https://app.aave.com) → Supply | - Network accessline 239
|--------|-----| | Supply collateral | [app.aave.com](https://app.aave.com) → Supply | | Find debt token | [app.aave.com](https://app.aave.com) → Reserve page → token dropdown, or [deployments.md](../deployments.md) | - Network accessline 240
| Supply collateral | [app.aave.com](https://app.aave.com) → Supply | - Network accessline 241
| Find debt token | [app.aave.com](https://app.aave.com) → Reserve page → token dropdown, or [deployments.md](../deployments.md) | - Network accessline 243
| Revoke delegation | Block explorer → VariableDebtToken → `approveDelegation(agent, 0)` | | Check delegation | `./aave-status.sh` or block explorer → `borrowAllowance(you, agent)` | | Monitor health | [app.aave.com](https://app.aave.com) d - Network accessline 244
| Check delegation | `./aave-status.sh` or block explorer → `borrowAllowance(you, agent)` | | Monitor health | [app.aave.com](https://app.aave.com) dashboard | | Repay debt | [app.aave.com](https://app.aave.com) → Repay, or agent runs `aave - Network accessline 245
| Monitor health | [app.aave.com](https://app.aave.com) dashboard | - Network accessline 246
| Repay debt | [app.aave.com](https://app.aave.com) → Repay, or agent runs `aave-repay.sh` | - Network accessline 264
Built by [Aaron Elijah Mars](https://aaronjmars.com), founder of Aeon and MiroShark · [@aaronjmars](https://github.com/aaronjmars)
.github/SECURITY.md
- Network accessline 17
➡<U+FE0F> **[Report a vulnerability](https://github.com/aaronjmars/agent-credit/security/advisories/new)** - Network accessline 49
Security fixes land on the `main` branch of [`aaronjmars/agent-credit`](https://github.com/aaronjmars/agent-credit). - Network accessline 50
[`aaronjmars/agent-credit`](https://github.com/aaronjmars/agent-credit).
.github/workflows/ci.yml
SKILL.md
- Network accessline 16
- **[OpenClaw](https://openclaw.ai/)** — Install as a skill, the agent borrows autonomously - Network accessline 17
- **[Claude Code](https://www.npmjs.com/package/@anthropic-ai/claude-code)** — Run scripts directly from a Claude Code session - Network accessline 20
Combines with **[Bankr](https://bankr.bot/)** skills for borrow-then-swap flows: borrow USDC via delegation, then use Bankr to swap, bridge, or deploy it. - Dangerous shell operationline 74
curl -L https://foundry.paradigm.xyz | bash && foundryup - Network accessline 74
curl -L https://foundry.paradigm.xyz | bash && foundryup - Network accessline 86
{ "chain": "base", "rpcUrl": "https://mainnet.base.org", - Network accessline 87
"chain": "base", "rpcUrl": "https://mainnet.base.org", "agentPrivateKey": "0xYOUR_AGENT_PRIVATE_KEY", - Network accessline 88
"rpcUrl": "https://mainnet.base.org", - Network accessline 321
- **Aave V3 Docs**: https://docs.aave.com/developers - Network accessline 322
- **Credit Delegation Guide**: https://docs.aave.com/developers/guides/credit-delegation - Network accessline 323
- **Aave Address Book**: https://github.com/bgd-labs/aave-address-book - Network accessline 324
- **Foundry Book**: https://book.getfoundry.sh/ - Network accessline 325
- **DebtToken Reference**: https://docs.aave.com/developers/tokens/debttoken
aave-borrow.sh
aave-repay.sh
aave-setup.sh
- Executable scriptline 1
#!/usr/bin/env bash - Access to secretsline 53
[Credential-related source omitted] - Access to secretsline 54
[Credential-related source omitted] - Access to secretsline 104
[Credential-related source omitted] - Access to secretsline 105
[Credential-related source omitted] - Access to secretsline 137
[Credential-related source omitted] - Access to secretsline 138
[Credential-related source omitted] - Access to secretsline 157
[Credential-related source omitted] - Access to secretsline 162
[Credential-related source omitted] - Access to secretsline 176
[Credential-related source omitted]
aave-status.sh
- Executable scriptline 1
#!/usr/bin/env bash - Access to secretsline 128
[Credential-related source omitted] - Access to secretsline 139
[Credential-related source omitted] - Access to secretsline 162
[Credential-related source omitted] - Access to secretsline 163
[Credential-related source omitted] - Access to secretsline 169
[Credential-related source omitted] - Access to secretsline 170
[Credential-related source omitted]
config.example.json
contracts.md
- Network accessline 74
The delegator sets up collateral and delegation from **their own wallet** — via [app.aave.com](https://app.aave.com) and a block explorer. See [README.md](.github/README.md) for the full step-by-step guide. - Network accessline 92
- **Aave Address Book (npm/Solidity)**: https://github.com/bgd-labs/aave-address-book - Network accessline 93
- **Deployed Contracts Docs**: https://docs.aave.com/developers/deployed-contracts - Network accessline 94
- **Interactive Dashboard**: https://aave.com/docs/resources/addresses
deployments.md
- Network accessline 3
Cross-check against [aave-address-book](https://github.com/bgd-labs/aave-address-book) before use. The credit delegation functions (`approveDelegation`, `borrowAllowance`, `borrow`, `repay`) use identical signatures on both V2 and V3 — but - Network accessline 147
- [Aave V3 Deployed Contracts](https://docs.aave.com/developers/deployed-contracts) - Network accessline 148
- [Aave V2 Deployed Contracts](https://docs.aave.com/developers/v/2.0/deployed-contracts) - Network accessline 149
- [bgd-labs/aave-address-book](https://github.com/bgd-labs/aave-address-book)
docs/assets/btn-docs.svg
- Network accessline 1
<svg xmlns="http://www.w3.org/2000/svg" width="110" height="40" viewBox="0 0 110 40" role="img" aria-label="Docs">
docs/assets/btn-star.svg
- Network accessline 1
<svg xmlns="http://www.w3.org/2000/svg" width="96" height="40" viewBox="0 0 96 40" role="img" aria-label="Star">
docs/assets/btn-x.svg
- Network accessline 1
<svg xmlns="http://www.w3.org/2000/svg" width="164" height="40" viewBox="0 0 164 40" role="img" aria-label="@aaronjmars on X">
docs/assets/hero-animated.svg
- Network accessline 1
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 1600 480" width="1600" height="480" role="img" aria-label="AGENT CREDIT — credit delegation for AI agents. Give your agent a credit line: it borrows from Aave when it needs funds, you sta
lib.sh
safety.md
tests/mocks/cast
- Executable scriptline 1
#!/usr/bin/env bash
tests/test-borrow-contract.sh
- Executable scriptline 1
#!/usr/bin/env bash - Dangerous shell operationline 37
trap 'rm -rf "$WORK_DIR"' EXIT - Network accessline 41
{ "chain": "test", "rpcUrl": "http://127.0.0.1:0", - Network accessline 42
"chain": "test", "rpcUrl": "http://127.0.0.1:0", "agentPrivateKey": "0x0000000000000000000000000000000000000000000000000000000000000001", - Network accessline 43
"rpcUrl": "http://127.0.0.1:0", - Dangerous shell operationline 193
rm -f "$WORK_DIR/send.log"
tests/test-repay-contract.sh
- Executable scriptline 1
#!/usr/bin/env bash - Dangerous shell operationline 34
trap 'rm -rf "$WORK_DIR"' EXIT - Network accessline 38
{ "chain": "test", "rpcUrl": "http://127.0.0.1:0", - Network accessline 39
"chain": "test", "rpcUrl": "http://127.0.0.1:0", "agentPrivateKey": "0x0000000000000000000000000000000000000000000000000000000000000001", - Network accessline 40
"rpcUrl": "http://127.0.0.1:0", - Dangerous shell operationline 134
rm -f "$WORK_DIR/send.log"
How the signals fit together
Facts from a static read of the files at the pinned commit. They describe what the code does, not why.
Some files could not be read; the facts below still hold.
Combined signals: 1 fact
- Passes content downloaded from foundry.paradigm.xyz to an interpreterSKILL.md:74High confidence
Signal confidence: 101 facts
- Network access in prose of documentation.github/CONTRIBUTING.md:21Low confidence
- Network access in prose of documentation.github/CONTRIBUTING.md:22Low confidence
- Network access in prose of documentation.github/CONTRIBUTING.md:23Low confidence
- Network access in prose of documentation.github/CONTRIBUTING.md:34Low confidence
- Network access in code of documentation.github/CONTRIBUTING.md:38Low confidence
- Network access in prose of documentation.github/README.md:5Low confidence
- Network access in prose of documentation.github/README.md:6Low confidence
- Network access in prose of documentation.github/README.md:7Low confidence
- Network access in prose of documentation.github/README.md:8Low confidence
- Network access in prose of documentation.github/README.md:9Low confidence
- Network access in prose of documentation.github/README.md:24Low confidence
- Network access in prose of documentation.github/README.md:25Low confidence
- Network access in prose of documentation.github/README.md:26Low confidence
- Network access in prose of documentation.github/README.md:27Low confidence
- Network access in prose of documentation.github/README.md:33Low confidence
- Network access in prose of documentation.github/README.md:34Low confidence
- Network access in prose of documentation.github/README.md:37Low confidence
- Dangerous shell operation in code of documentation.github/README.md:91Low confidence
- Network access in code of documentation.github/README.md:91Low confidence
- Access to secrets in prose of documentation.github/README.md:102Low confidence
- Network access in prose of documentation.github/README.md:133Low confidence
- Network access in prose of documentation.github/README.md:143Low confidence
- Network access in prose of documentation.github/README.md:215Low confidence
- Network access in prose of documentation.github/README.md:219Low confidence
- Network access in prose of documentation.github/README.md:225Low confidence
- Network access in prose of documentation.github/README.md:226Low confidence
- Network access in prose of documentation.github/README.md:227Low confidence
- Network access in prose of documentation.github/README.md:238Low confidence
- Network access in prose of documentation.github/README.md:239Low confidence
- Network access in prose of documentation.github/README.md:240Low confidence
- Network access in prose of documentation.github/README.md:241Low confidence
- Network access in prose of documentation.github/README.md:243Low confidence
- Network access in prose of documentation.github/README.md:244Low confidence
- Network access in prose of documentation.github/README.md:245Low confidence
- Network access in prose of documentation.github/README.md:246Low confidence
- Network access in prose of documentation.github/README.md:264Low confidence
- Network access in prose of documentation.github/SECURITY.md:17Low confidence
- Network access in prose of documentation.github/SECURITY.md:49Low confidence
- Network access in prose of documentation.github/SECURITY.md:50Low confidence
- Network access in prose of the SKILL.mdSKILL.md:16Low confidence
- Network access in prose of the SKILL.mdSKILL.md:17Low confidence
- Network access in prose of the SKILL.mdSKILL.md:20Low confidence
- Dangerous shell operation in code of the SKILL.mdSKILL.md:74High confidence
- Network access in code of the SKILL.mdSKILL.md:74High confidence
- Network access in code of the SKILL.mdSKILL.md:86High confidence
- Network access in code of the SKILL.mdSKILL.md:87High confidence
- Network access in code of the SKILL.mdSKILL.md:88High confidence
- Network access in prose of the SKILL.mdSKILL.md:321Low confidence
- Network access in prose of the SKILL.mdSKILL.md:322Low confidence
- Network access in prose of the SKILL.mdSKILL.md:323Low confidence
and 51 more
Works with
| Agent | Status | Evidence | Install to |
|---|---|---|---|
| Claude Code | Compatible | Standard SKILL.md SKILL.md | .claude/skills/<skill-name>/ |
| Codex | Compatible | Standard SKILL.md SKILL.md | .agents/skills/<skill-name>/ |
| Cursor | Compatible | Standard SKILL.md SKILL.md | .cursor/skills/<skill-name>/ |
| Gemini CLI | Compatible | Standard SKILL.md SKILL.md | .gemini/skills/<skill-name>/ |
| GitHub Copilot | Compatible | Standard SKILL.md SKILL.md | .github/skills/<skill-name>/ |
| OpenCode | Compatible | Standard SKILL.md SKILL.md | .opencode/skills/<skill-name>/ |
- Native The repository ships a manifest, folder or frontmatter field for this agent.
- Compatible A standard SKILL.md that this agent can read.
- Unknown No evidence either way.
Derived automatically from paths, manifests and frontmatter. Not tested on each agent.