inkbox-ai/openclaw-plugin/inkbox-credential-use
Claude Code and agent skill by inkbox-ai/openclaw-plugin
- community
Checking that the SKILL.md file is still in the repository.
Use when the user asks the agent to "log into X", "get the API key for Y", "fetch the SSH key for Z", or "give me the TOTP code for service A". Covers the credential vault path through the Inkbox plugin. Always list-then-get; never enumerate plaintext.
Install
npx vskill@latest install inkbox-ai/openclaw-plugin/inkbox-credential-useSKILL.md
Inkbox credential use
When an action requires plaintext credentials (a password, an API key, an SSH key, a TOTP code), use this skill to retrieve them through the Inkbox vault. The vault tools are all optional — the user must opt them into tools.allow before this skill can do anything.
Required tools (all optional, must be allowlisted)
inkbox_credentials_list— metadata only (id, name, secretType). Safe to call freely.inkbox_credentials_get_login— plaintext login (username + password + optional URL)inkbox_credentials_get_api_key— plaintext API keyinkbox_credentials_get_ssh_key— plaintext SSH private key (+ public, fingerprint, passphrase)inkbox_credentials_get_secret— one selected secret, including key pairs/other; optional separatelyinkbox_totp_code— current TOTP code for a login that has TOTP configured
Prerequisites
- The vault must be initialized in the Inkbox Console.
- The vault unlock key must be available in the
INKBOX_OPENCLAW_VAULT_KEYenv var (or a custom env var ifvault.keyEnvVaris configured). - The identity must have access grants to the secrets in question (admin-set via the Inkbox Console).
Metadata listing works while locked. Reads refresh access and payloads on every call; login results expose has_totp, never a seed. If INKBOX_OPENCLAW_VAULT_KEY is not set, a plaintext/code tool call returns "Vault is locked." Direct the user to export the env var in the shell launching OpenClaw, then retry.
Workflow
- List first. Call
inkbox_credentials_list(optionally filter bytype). Read the metadata to find the rightidbyname. Never guess UUIDs.
- Confirm with the user when stakes are high. If the user said "log into AWS production," verify the matching secret's
nameanddescriptionlook right before fetching plaintext. The agent should not silently pull production credentials.
- Fetch the typed plaintext. Call the matching get tool:
- Login →
inkbox_credentials_get_login - API key →
inkbox_credentials_get_api_key - SSH key →
inkbox_credentials_get_ssh_key
Each returns the payload as JSON. Use the fields directly in the action the user asked for; do not echo the plaintext back to the user unless they explicitly asked to see it.
- TOTP is a separate flow. Call
inkbox_totp_codewith the login secret's UUID — returns{ code, secondsRemaining }. The code is valid forsecondsRemainingseconds; if it's <5, regenerate before using.
Hygiene
- Don't list-then-dump.
credentials_listreturns metadata only by design. Don't call all threeget_*tools in a loop trying to enumerate the vault. - Don't paste plaintext into chat. When the user asks "use the GitHub PAT to push," fetch it and use it — don't repeat it back to them.
- Don't store outside this call. The plugin caches the unlocked vault in-process; you don't need to (and shouldn't) keep plaintext in session memory.
Errors
| Error | Meaning | |
|---|---|---|
Vault is locked. Set the INKBOX_OPENCLAW_VAULT_KEY... | Env var missing — direct the user to set it. | |
404 on get_* | Wrong secret id, or this identity doesn't have access. | |
TypeError on get_login/get_api_key/get_ssh_key | Caller picked the wrong typed getter for the secret's type. Re-list to see secretType. |
What this skill does NOT cover
- Creating, updating, or deleting secrets — there's no plugin tool for this in agent-scoped mode.
- Granting access to secrets — admin-only via the Inkbox Console.
- TOTP setup — initial TOTP config also happens in the Console.
When you need more — raw Inkbox docs
If a payload shape, secret type, vault behavior, or TOTP detail isn't covered here, go to the source:
- https://inkbox.ai/llms.txt — LLM-friendly index of every Inkbox doc page.
- https://inkbox.ai/docs/all.md — the full Inkbox documentation concatenated as one markdown file.
Especially useful when checking the exact fields on LoginPayload, APIKeyPayload, SSHKeyPayload, KeyPairPayload, OtherPayload, or TOTPCode rather than guessing.
Security
- Security scanPassed, 90 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingskills/inkbox-credential-use/SKILL.md
Popularity
- 7-day trend score
- 0
Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/inkbox-ai/openclaw-plugin/inkbox-credential-use)