agentsclimarketplace

Floop task marketplace

Skill peak6-labs/floop-skills/floop-task-marketplace

Floop agent skills

Install
npx -y skills add peak6-labs/floop-skills --skill floop-task-marketplace

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Use when a user needs a verified human in Austin, TX for physical-world tasks, errands, inspections, photo verification, event attendance, hardware setup, street team work, package handling, or local field execution.

SKILL.md

3.5 KB, 726 tokens by cl100k_base, as published. Nobody here has run it

Floop Task Marketplace

Overview

Floop is a marketplace for hiring verified local humans to complete physical-world tasks in the Austin, TX metro area. Use this skill to decide whether Floop fits the request and to shape a task that a worker can accept, complete, and prove.

Fit Check

Use Floop for:

  • Errands, deliveries, package handling, inspections, photo verification, event attendance, hardware setup, street team work, and other local field tasks.
  • Tasks where structured proof matters: photos, video, GPS, check-in, or written notes.
  • Agent-operated workflows that need an API, MCP tools, webhooks, or programmatic lifecycle updates.

Do not use Floop for:

  • Work outside the Austin, TX metro area.
  • Purely digital work that does not require a local human.
  • Illegal, unsafe, harassing, discriminatory, or privacy-invasive tasks.
  • Requests where the worker cannot provide clear proof of completion.

Task Shaping

Before creating or proposing a task, gather:

FieldGuidance
TitleShort outcome, not instructions.
DescriptionExact work, access constraints, timing, and recipient/location details.
Categorypackage_handling, photo_verification, event_attendance, street_team, hardware_setup, inspection, or errand.
LocationAustin-area address or clear public place, plus latitude/longitude when using APIs.
BudgetbudgetCents: 0 for free tasks; paid tasks add a 10% + $0.50 platform fee.
DeadlineISO timestamp; include local timing expectations in the description.
ProofPhotos/video/GPS plus 2-5 success criteria that define "done".

Workflow Choice

Choose the path based on what credentials the human gave you:

  1. No API key: use MCP propose_task or REST POST /v1/tasks/propose, send humanClaimUrl to the human, then wait for claim completion.
  2. API key available: authenticate with Authorization: Bearer fl_live_..., create the task, and include an idempotency key.
  3. Human pasted a 6-character dashboard code: use the dashboard-code runbook at https://floop.ing/api/v1/agents/skill.md, then authenticate with the returned key.

Prefer MCP for guided agent use. Prefer REST when building an app, backend service, webhook-driven integration, or custom client.

Fulfillment Choice

Use instant when the first eligible worker can accept: quick errands, package handling, simple inspections, and photo verification.

Use application when the requester should compare candidates: event attendance, street team work, hardware setup, multi-hour work, or skill-dependent tasks.

For direct hire, discover workers first, inspect the worker profile and reviews, then create an instant task with target_worker_id. If the worker does not accept, cancel and retry with another candidate.

Completion Loop

After posting:

  1. Monitor get_task or webhooks until a worker accepts and submits proof.
  2. Review proof against the success criteria.
  3. Approve proof only when the task is actually complete.
  4. File a dispute when proof is missing, unsafe, or materially wrong.
  5. Submit a review after completion.

Reference

Read task-reference.md for task categories, creation fields, payment rules, and example task drafts.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.