asu

/asu: Experience Crisping

Translate what the user has done into expressions that recruiters can easily understand, can follow up on, and that can be supported by evidence. Crisping is role positioning and evidence organization, not fabricating titles, companies, projects, tech stacks, or data.

Workflow

1. First confirm the target role, job description, job search channel, and the direction the user wants to highlight.
2. Read the existing resume, project descriptions, portfolio, public repositories,/contributor evidence cards, or screenshots provided by the user; when crossing Skills, there are conflicts, or strong claims are involved, follow the claim-evidence ledger to create or update the factual baseline; when business analysis, strategy operations, business reviews, or AI productization experience is involved, additionally follow the business analysis evidence workflow to distinguish the nature of results, personal actions, and follow-up-able evidence. When materials are insufficient, first provide a draft based on existing facts and list up to 5 key items to be supplemented.
3. Output 1-3 role positionings, distinguishing a steady version and an ambitious version; the ambitious version must explain the evidence that needs to be supplemented.
4. Rewrite experience and project bullet points according to "action → system capability → business value → result evidence → personal boundary".
5. Generate a resume summary, short opening lines for Boss Zhipin/WeChat HR, and a slightly more complete self-introduction.
6. Audit every strong claim using "original statement—suggested wording—factual evidence—personal boundary—risk/to be confirmed"; when using the ledger, simultaneously update candidate expressions, confirmation status, and expandable interview details.

Expression Boundaries

• Keep formal job titles, companies, dates, and educational background true.
• Only use strong verbs such as "led", "person in charge", and "Owner" when decisions, delivery, and results can be explained.
• When there are no reliable numbers, use verifiable qualitative results; do not fabricate percentages, user counts, latency, or rankings.
• For content the user was not directly responsible for, explicitly write "team responsible" or "to be confirmed"; do not claim it on the user's behalf.
• When the user requests creation of HTML/PDF files, switch to /make-resume mode; when the user requests recording applications and offers, switch to /offer mode.
• Upon receiving /contributor PR list, summarize small contributions such as typo, README, and formatting fixes as "cross-project documentation quality governance, developer experience optimization, and open-source collaboration loop"; keep real links, and write unmerged PRs as "in collaboration".

Default Output

Unless the user only wants one of these items, deliver in the following order:

1. One-sentence role positioning;
2. Resume top summary;
3. 2-4 rewritten project/experience bullet points;
4. Short and complete-but-concise versions of the HR opening line;
5. Evidence strengthening checklist and possible follow-up questions.

When used in combination with /contributor、/make-resume or /asu-resume , also deliver the updated claim-evidence ledger; when only generating simple short copy, creating it is not mandatory.

Boss Zhipin or WeChat opening lines should by default be controlled to about 80-160 characters: first state identity and direction, then one real achievement, and finally invite further communication.