oa-prep
Skill2 starsChanged 26 days agoMITView on GitHub
Build or update an online-assessment (OA) and interview prep guide for one company from your OAHelper Pro account (the `oahelper` MCP server) plus public sources. Use when the user says "prep me for X", "make a prep guide for X", "what does X ask in the OA", "X OA questions", or names a company with an upcoming OA or interview.
In this repository8 other files it gives its agents.
- interview-insightsSkill
- mcq-drillSkill
- oa-compareSkill
- oahelper-setupSkill
- oa-mockSkill
- revise-mistakesSkill
- solve-coachSkill
- study-planSkill
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 1 is distinctive, and 3 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 26
Read the pack: Start with library/<co>/index.md, then the sets, questions and interviews.
- Very specific: line 34
If a guide already exists, add a ## From OAHelper (<date>) section instead of a second file.
- Distinctive: line 42
Every past OA question: one line with its set name and date, the topic, the closest known LeetCode or CSES problem, and a link to the saved …
- Very specific: line 50
Write unknown values as UNKNOWN.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
OA prep from OAHelper
Part of oahelper-mcp by Sirjan Singh.
OAHelper content comes from the oahelper MCP server. Its library at
library/<company>/ in the data folder (oahelper_status prints where) is the offline source of truth: read the files there
rather than re-fetching.
OAHelper's terms forbid bots and redistribution, and the penalty can be losing Pro. Fetch one company at a time, only when asked. Never copy library content into anything that leaves this machine (posts, GitHub, shared docs).
Steps
- Offline first.
oahelper_offline(company). If the pack exists and is under a week old, go to step 3. - Sync.
oahelper_sync(company), repeated while it saysNOT DONE, capped at about 60 live requests for one guide. If the search shows the company under a differentunique_company_name, sync that name. Leaveinclude_imagesoff unless a statement says its screenshots carry the only copy of the input or the options. - Read the pack. Start with
library/<co>/index.md, then the sets, questions and interviews. Weight by date: the last 12 months decide the format; older sets are pattern evidence only. Put the year on every item you cite. - Add public sources: LeetCode Discuss, GeeksforGeeks interview experiences, the company's own careers pages. Search before asserting anything about the hiring process.
- Official solutions (
oahelper_solution) only for the few coding questions where the approach is not obvious. Solve the rest yourself and label them as your own. - Write the guide to
prep/<co>-oa-prep.md(or wherever the user keeps prep notes). If a guide already exists, add a## From OAHelper (<date>)section instead of a second file.
Guide format
- Header: role, OA date and time if known, how the guide was compiled and when.
## What the evidence supports: a table of the newest reports (date, setting, format, source), then the converged picture: question count, time limit, scoring, topic mix.- Every past OA question: one line with its set name and date, the topic, the closest known LeetCode or CSES problem, and a link to the saved file.
- MCQ section: recurring themes from the answer keys, with a few representative questions and their answers.
- Interview rounds from the experience write-ups: what each round tests, what sank people.
- A time-boxed study plan when the OA date is known.
## Sources: OAHelper URLs from the file headers, plus public links.
Write unknown values as UNKNOWN; never guess.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

