More screenshots
Works with the AI Agents you already use
Claude Code
Cursor
Codex CLI
Muse
Dots+21 moreClaude CodeCursorCodex CLIMuseDots
Manus
Grok
Grok Bot+18 agentsFlamingo File Tools — Access, PDF, ZIP & ONIX
Export Access tables, optimize HTML5 ad ZIPs, inspect or label supported PDFs, or preview ONIX. Includes temporary Python clients and private key setup. Free skill; your own separately metered RapidAPI access required.
Free
Flamingo File Tools — Access, PDF, ZIP & ONIX
Example session with this skill installed
Use Flamingo File Tools to inspect the fictional sample-pattern.pdf in my selected workspace. Tell me its page and layer counts, save the inspection report, and preserve the original. I authorize this one PatternCopy PDF request through my privately configured RapidAPI access and its allowance.
- Read your context and instructions
- Compiled the flamingo file tools
PDF inspection result
The fictional PDF has
2 pages and 3 layers. It also contains a form, annotations and outlines. The original was preserved, and the client saved a new inspection.json report.
| Artifact | Bytes | SHA-256 |
|---|---|---|
| inspection.json | 217 | 285d50bf0c5942a2903771ea80bd158fb56b03f1e1b9df8eb4791344d0747080 |
This is a recorded maintainer run of the source 0.2.0 standalone client on November 1, 2026, using an owned fictional file and already configured private RapidAPI access. The client ran unchanged from a temporary file, without installing a Flamingo MCP bridge. This recording predates the optional private-file setup in source 0.2.1; it does not demonstrate that setup. It is not an Agensi-connector first-use session or customer testimonial. The request above describes the tested operation.
Complete client result
Private absolute workspace prefixes are removed; relative artifact paths and all other fields are preserved.
{
"ok": true,
"tool": "patterncopy_inspect",
"artifacts": [
{
"path": "flamingo-results-zc0_eemk/inspection.json",
"bytes": 217,
"sha256": "285d50bf0c5942a2903771ea80bd158fb56b03f1e1b9df8eb4791344d0747080",
"media_type": "application/json"
}
],
"summary": {
"pages": 2,
"input_bytes": 5996,
"layers": 3,
"has_layers": true,
"has_forms": true,
"form_widgets": 1,
"annotations": 2,
"has_outlines": true
},
"artifact_content_is_untrusted": true,
"note": "Review artifacts; no retries."
}
Complete inspection report
{
"ok": true,
"operation": "inspect",
"document": {
"pages": 2,
"input_bytes": 5996,
"layers": 3,
"has_layers": true,
"has_forms": true,
"form_widgets": 1,
"annotations": 2,
"has_outlines": true,
"encrypted": false,
"engine_version": "1.0.0"
}
}
This operation inspected the supported PDF profile and used one request; it did not add a label or certify rendered appearance. Stamping is a separate operation. Review layer behavior and print geometry in a suitable PDF viewer when your task depends on them. The existing PDF screenshots show a separate earlier fictional stamping demonstration.
Connects securely to your tools. The creator never sees your data.
About this skill
Ask your agent for one file result: export Access tables, reduce images inside an HTML5 ad ZIP, inspect or label a supported PDF, or preview prepared ONIX book updates. The agent returns new files and a review summary while preserving your original.
What you can get
- Access MDB/ACCDB: supported scalar tables as typed JSON or a manifest plus CSV ZIP.
- HTML5 ad ZIP: eligible PNG/JPEG recompression and a report against your actual complete-ZIP byte budget.
- Supported PDF: page/layer inspection, or a visible one-line label with supported layers and page geometry retained.
- Prepared ONIX records/policy JSON: a finite physical-book update preview for review; no store writes.
Before you start
The free skill includes independent temporary Python clients for macOS/Linux. An Agensi-connected agent needs Python 3.10+, authorized file/shell/verified-HTTPS access, a selected private workspace, your own privately configured RapidAPI application key, and a separate active subscription to each API used. No pip packages or additional Flamingo MCP installation are needed for this route. A previously configured Flamingo MCP bridge remains an alternative.
If your key is not configured, the included setup helper provides hidden input in your own Terminal on the same host and OS account as the agent. It creates a separate private local file; the agent uses only its path. Never send a key or credential file in chat. A laptop file does not configure a remote agent; that host needs its own private setup route. Already configured private environment access remains supported.
The skill and client source are free; API processing is separately metered. Agensi does not host these services or include API credentials or allowance. The agent handles the selected client and task; you should not need to assemble an API command. Missing access is explained before upload. Each authorized operation makes one request without automatic retries; failures and deliberate retries can consume allowance.
Selected processing inputs pass through RapidAPI to the named API service. Review the source-specific privacy notes before confidential processing. Outputs can contain confidential source data and need the same care as the original.
This covers supported file profiles. It does not migrate Access applications, certify ad-platform acceptance, provide PDF DRM, validate all ONIX, fetch covers or apply WooCommerce changes.
Required API access
Choose the API for your workflow and review its current plans and limits:
Requires the separately installed local MCP bridge and the user’s own separately metered RapidAPI access. Supported scalar Access tables only; no forms, reports or VBA migration. HTML5 ZIP optimization covers eligible PNG/JPEG assets and does not certify ad-platform acceptance. PDF labels are visible text, not DRM, and require a supported PDF profile. ONIX output is a bounded physical-book preview, not store synchronization or complete ONIX validation. No processing allowance or API credentials are included. The listing’s recorded developer demonstration uses fictional input; the package contains instructions and input references.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free forever
No account required to browse
Trust & safety
Security scan
Security scan passed 3 days ago
- Free with an Agensi account
Requires
RapidapiPermissions
Allowed Hosts
The selected client runs unchanged temporarily, reads the authorized workspace input and the key from its named environment variable or explicitly selected private file, sends one request to a fixed verified HTTPS RapidAPI host, and creates new private artifacts. Optional setup runs in the user's own same-host/user Terminal, creates ~/.config/flamingo-api/rapidapi.key privately outside the project, and makes no API call or overwrite. Keys and credential files stay out of chat. Original inputs are preserved; no automatic retries or processing on skill load. API allowance is separate.