OpenClaw researches apprenticeships into Notion
Gangue reports that OpenClaw produced working apprenticeship-offer links in Notion after eight hours of joint debugging, with some job sites inaccessible.
borisgangue.dev
Suggested prompt
Suggested prompt written by Agents In Use, not the source author. Replace bracketed inputs and review tool permissions first.
Find apprenticeship opportunities matching [criteria] and store source links in [Notion database]. Use current API documentation, report inaccessible sites, and ask me before sending any application or personal information.
Reach people exploring AI agents
See placements →Evidence at a glance
Sourced account- Source author
- Boris Gangue
- Original account
- Read the source ↗
- Published
- Source link checked
- Reported result
- Gangue reports that OpenClaw produced working apprenticeship-offer links in Notion after eight hours of joint debugging, with some job sites inaccessible.
- Human involvement
- Gangue provides API keys, supplies documentation information and debugs with the assistant. Application submission and successful hiring are not confirmed.
A linked account supports attribution. It does not mean we independently reproduced its result.
Report an issue with this case
Open correction emailOpens your email app with the case link included. Nothing is sent automatically.
How the agent did it
- Run OpenClaw on a spare desktop and connect Discord.
- Give the assistant its own email account and Notion access.
- Ask for an apprenticeship-search workflow and review its plan.
Boris Gangue configured an OpenClaw assistant on an old HP desktop and used Discord as its interface. The agent began with Gemini 2.5 Flash. After a basic chat trial, he gave it its own email account and access to Notion, then asked for an automated system to find and apply for apprenticeship opportunities.
The assistant proposed a plan, but implementation required roughly eight hours of debugging with Gangue. Notion’s changed API was a major obstacle. At that point the agent lacked a web-search tool, so Gangue supplied the information it needed and provided API keys himself. He reports that the final output contained working links to opportunities, while the quality remained below what he had hoped.
The post does not establish that applications were actually submitted or that an apprenticeship was secured. It also reports the assistant’s explanation that some EU job platforms could not be scraped or accessed; that explanation is not an independently checked legal conclusion. The transferable workflow is to separate opportunity research and saved records from application submission, supply current integration documentation, and review the resulting links. Gangue recommends isolating the assistant on a spare machine or virtual machine because of its broad system access.
Prompt
Suggested prompt (written by Agents In Use, not the author's)
Source: Blog · Boris Gangue
Frequently asked questions
Did the author obtain an apprenticeship?
The post reports working opportunity links, not a successful application or hiring outcome.
Why did implementation take so long?
Gangue describes Notion API changes and the agent’s lack of web-search access during the initial setup.
How was the assistant isolated?
He ran it on an old HP desktop and recommends a separate machine or virtual machine for experiments.
More details
Remaining steps
- Supply updated Notion API information during debugging.
- Check the opportunity links in the resulting output.
About this account
OpenClaw plans and implements the search/integration workflow, but lacked web search during the Notion debugging phase and needed the user’s help.
Firsthand account by Boris Gangue; original publication date verified against Primary page publication date. Outcomes are the author’s reports, not independently reproduced. Suggested prompt written by Agents In Use; not the author’s original prompt.
Credit: Boris Gangue
Source published
Added to Agents In Use
Source checked
Tools and topics
Discord · Notion · HP ProDesk · Gemini 2.5 Flash
Job research · Notion · Integration debugging
Editorial completeness: 100/100. Measures documented fields, not agent performance or verified outcomes.