# Company Research > Company Research - Skill: `handsomeboy990/company-research` (Agent Skill, multi-file: 4 files) - Install (CLI): `npx skillmds@latest add handsomeboy990/company-research` - Raw SKILL.md: https://api.skillmd.com/api/skills/handsomeboy990/company-research/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Research & Search - Author: Handsomeboy990 (https://skillmd.com/u/handsomeboy990) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/handsomeboy990/company-research --- # Company Research A candidate applies to and joins a company, not a job posting. Company research tells them what they are actually walking into: what the company does, whether it is healthy, what it is like to work there, and what to ask. Every fact is from a live source, because a candidate makes a real decision on it, and an invented fact about an employer is worse than none. ## 1. What the company actually does, and how it earns ``` product what the company makes or does, in plain terms, from its own site and product, not its marketing tagline model how it actually makes money: who pays, for what; a company whose revenue model is unclear is itself a signal market who its customers and competitors are, from real sources stage startup, scale-up, or established; each is a different bet for the candidate, with different risks and rewards ``` ## 2. Health signals from real evidence ``` funding for a startup: the last raise, when, how much, from whom, from real announcements; a long-ago raise with heavy hiring is a runway question growth hiring trends, product launches, customer announcements, from live sources; expansion and contraction both show stability layoffs, leadership churn, public trouble, from reporting; named if found, not speculated if not caveat these are signals, not verdicts; a candidate weighs them, and the research presents them with their source and their date, not as a prediction ``` ## 3. Culture and reputation, from balanced sources ``` sources employee reviews, the company's own statements, reporting, and where possible people the candidate can actually talk to; each weighed by its bias balance review sites skew to the aggrieved and the coached-positive; read the pattern across many, not one dramatic entry specific "several reviews mention long hours during releases" is useful; "the culture is bad" is not, and neither is repeating one anonymous rant signal consistent themes across independent sources carry weight; a single unverifiable claim is a question to ask, not a fact to report ``` ## 4. The role's real context ``` team where the role sits, who it reports to, the team's size and remit, from the listing and any reachable source why open is it growth (new headcount) or backfill (someone left), and what that suggests; a role open for many months is a question stack the technologies and tools the role really uses, for the technical interview prep challenge the real problem the role exists to solve, which the cover letter and interview both use ``` ## 5. Red flags a candidate should weigh ``` name the signals worth a candidate's attention: unclear revenue with a long-past raise, high leadership churn, a role reposted repeatedly, a review pattern of the same specific complaint, a hiring process that itself behaves badly frame as a weighted concern with its source, not a verdict; the candidate decides, and some flags are acceptable for the right role verify distinguish a sourced pattern from a single anonymous claim; report the first as a concern and the second as a question to raise ``` ## 6. Prohibitions - Never invent a fact about a company: a funding round, a revenue figure, a headcount, a customer, a piece of news. - Never report an unsourced rumour as a fact; mark it as a question to verify. - Never repeat a single dramatic review as the company's culture; read the pattern. - Never present a signal as a prediction; a candidate weighs signals, they are not verdicts. - Never let the candidate's hope for the role suppress a real red flag. - Never speculate about trouble not found; absence of a signal is stated as such. ## 7. Protocol 1. Establish what the company does and how it earns, from its own sources. 2. Gather health signals from real, dated evidence; present with source, not prediction. 3. Assess culture from balanced sources, reading patterns not single entries. 4. Establish the role's real context: team, reason open, stack, challenge. 5. Name the red flags as weighted, sourced concerns, distinguished from single rumours. 6. Produce the profile the candidate uses to decide, tailor, and prepare. 7. Where a live source cannot be reached, say so and name what the candidate should check themselves. ## 8. Auto-critique Score from 0 to 5: what the company does and how it earns established from real sources, health signals dated and sourced not predicted, culture read as a pattern across balanced sources, role context real, red flags weighted and sourced, single rumours kept distinct from patterns, nothing invented. Threshold: no axis below 3, average at least 4. Any invented company fact, or a single anonymous review reported as the culture, caps the score until fixed, per career-core. ## 9. Interfaces - Upstream: `career-core` for the source and honesty rules. - Downstream: `cover-letter` uses the real reason and register, `interview- preparation` uses the role context and the candidate's questions, `job-search` uses the health assessment to rank a shortlist. - Lateral: `source-research` and `source-verification` in `research/` for the gathering and checking discipline; `competitive-analysis` when comparing several prospective employers. ## 10. Live sources Company research reads live sources: the company's own site and product, funding and news databases, employee-review sites, and any connector the runtime exposes. It presents each fact with its source and date. When a source cannot be reached, it names the gap and tells the candidate exactly what to look up themselves, rather than filling the company's profile with plausible invention, because the candidate is about to make a real decision on it.