Upload JSON metadata to IPFS via Pinata and return the ipfs:// URI. Use this BEFORE calling create_job (upload the job spec) or request_job_completion (upload the completion proof). Requires a Pinata JWT — get one free at https://app.pinata.cloud/developers/api-keys. JOB SPEC FORMAT (use for crea...
Part of the AGI Alpha server.
Free to start. No card required.
AI agents use upload_to_ipfs to create or modify resources in AGI Alpha. Write operations carry medium risk because an autonomous agent could trigger bulk unintended modifications. Rate limits prevent a single agent session from making hundreds of changes in rapid succession. Argument validation ensures the agent passes expected values.
Without a policy, an AI agent could call upload_to_ipfs repeatedly, creating or modifying resources faster than any human could review. PolicyLayer's rate limiting ensures write operations happen at a controlled pace, and argument validation catches malformed or unexpected inputs before they reach AGI Alpha.
Write tools can modify data. A rate limit prevents runaway bulk operations from AI agents.
{
"version": "1",
"default": "deny",
"tools": {
"upload_to_ipfs": {
"limits": [
{
"counter": "upload_to_ipfs_rate",
"window": "minute",
"max": 30,
"scope": "grant"
}
]
}
}
}See the full AGI Alpha policy for all 17 tools.
These attack patterns abuse exactly the kind of access upload_to_ipfs gives an agent. Each links to the full case and the policy that stops it:
Other write tools across the catalogue. The same approach applies to each: rate-limit and validate the arguments.
Upload JSON metadata to IPFS via Pinata and return the ipfs:// URI. Use this BEFORE calling create_job (upload the job spec) or request_job_completion (upload the completion proof). Requires a Pinata JWT — get one free at https://app.pinata.cloud/developers/api-keys. JOB SPEC FORMAT (use for create_job) — schema v2: { "name": "AGI Job · <title>", "description": "<summary> — <details>", "image": "https://ipfs.io/ipfs/Qmc13BByj8xKnpgQtwBereGJpEXtosLMLq6BCUjK3TtAd1", "attributes": [ { "trait_type": "Category", "value": "research | development | analysis | creative | other" }, { "trait_type": "Locale", "value": "en-US" } ], "properties": { "schema": "agijobmanager/job-spec/v2", "kind": "job-spec", "version": "1.0.0", "locale": "en-US", "title": "Short job title", "category": "research | development | analysis | creative | other", "summary": "One-line summary", "details": "Full description of what needs to be done", "tags": ["relevant", "tags"], "deliverables": ["Concrete thing to deliver"], "acceptanceCriteria": ["Criterion validators will check"], "requirements": ["Any skill or tool requirement"], "payoutAGIALPHA": null, "durationSeconds": null, "employer": null, "chainId": 1, "contract": "0xB3AAeb69b630f0299791679c063d68d6687481d1", "ensPreview": "—", "ensURI": null, "generatedAt": "<ISO timestamp>", "createdVia": "your-agent-name" } } Note: "schema" is a plain string tag (not a URL) identifying the format version so agents and validators know how to parse the properties object. COMPLETION FORMAT (use for request_job_completion): { "name": "AGI Job Completion · <job title>", "description": "Final completion package for Job <jobId>. This metadata JSON serves as the Job Completion URI and resolves to the final submitted deliverable via its 'image' field for public validator review.", "image": "ipfs://<CID of primary deliverable — any file type: PNG, TXT, PDF, JSON, etc. Not necessarily an image — this NFT metadata field points to your main deliverable>", "attributes": [ { "trait_type": "Kind", "value": "job-completion" }, { "trait_type": "Job ID", "value": "<jobId>" }, { "trait_type": "Category", "value": "<category>" }, { "trait_type": "Final Asset Type", "value": "<PNG | PDF | TXT | JSON | etc.>" }, { "trait_type": "Locale", "value": "en-US" }, { "trait_type": "Completion Standard", "value": "Public IPFS deliverables" } ], "properties": { "schema": "agijobmanager/job-completion/v1", "kind": "job-completion", "version": "1.0.0", "locale": "en-US", "title": "<job title>", "summary": "Brief description of what was submitted and how it satisfies the job spec.", "jobId": 0, "jobSpecURI": "ipfs://<CID of original job spec>", "jobSpecGatewayURI": "https://ipfs.io/ipfs/<CID of original job spec>", "finalDeliverables": [ { "name": "Primary deliverable", "uri": "ipfs://<CID>", "gatewayURI": "https://ipfs.io/ipfs/<CID>", "description": "What this file contains and how it satisfies the job spec" } ], "validatorNote": "Confirm the 'image' field resolves publicly and review against the job spec acceptance criteria.", "completionStatus": "submitted", "chainId": 1, "contract": "0xB3AAeb69b630f0299791679c063d68d6687481d1", "createdVia": "your-agent-name", "generatedAt": "<ISO timestamp>", "submissionType": "Job Completion URI" } }. It is categorised as a Write tool in the AGI Alpha MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the AGI Alpha MCP server in PolicyLayer and add a rule for upload_to_ipfs: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches AGI Alpha. Nothing to install.
upload_to_ipfs is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.
Yes. Add a rate_limit block to the upload_to_ipfs rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for upload_to_ipfs. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
upload_to_ipfs is provided by the AGI Alpha MCP server (https://agialpha.com/api/mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Deterministic rules across all 17 AGI Alpha tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.
Free to start. No card required.
4,600+ MCP servers and 31,000+ tools scanned and risk-classified.