A useful profile should help someone understand your capabilities. A learning space should help you develop them without pretending you know everything already.
A skill needs context
Show the discipline, project stages, environment, and contribution. Did you design, develop, deploy, operate, diagnose, review, or teach?
Illustrative skills: RAG evaluation · Agent tool integration · Multilingual speech recognition · Model quantization · AI security testing · Distributed-training diagnosis
01
Two different reviews
Catalog review: Is a proposed skill clear, meaningful, and distinct? Members can suggest missing skills, including rare capabilities that are not widely known. Human review comes before publication.
Professional review: What evidence supports an individual’s claim? A claim remains in the member’s workspace until the relevant review supports a scoped public label.
You can name a skill you want to learn without claiming proficiency. AI may help organize evidence and identify duplicates. People approve catalog entries and verifications.
Portfolios and reviews do different jobs
A portfolio presents authorized work. A review states what was checked. Make your role, collaborators, source permissions, and the actual review status clear.
Private mentoring feedback is not a public verdict. Publishing a portfolio does not independently verify every associated claim.
A professional badge says what was checked, by whom and when. Nothing more.
02
Recognition is not a qualification.
Flow Credits and founding or contribution recognition do not buy technical standing or a favorable review.
Recognition awards should be labeled separately from skill verification. Everyone’s professional evidence is assessed against the relevant standards, whether or not they hold a community award.
03
Private means private.
| Information |
Intended audience |
| Public profile, portfolio, and approved feedback |
The audience approved for publication. |
| Customer-private assessments, rates, and pool records |
Authorized people in that customer organization. |
| 3Fabrics.ai qualification notes |
People assigned to that qualification work. |
| Project-restricted evidence |
Explicitly authorized participants in that project. |
These boundaries apply to search, notifications, exports, and AI features. A public excerpt needs separate approval and must not expose private history.
Private designs will not be used to train AI models.
04
Fair feedback, not hidden blacklists.
Customers, experts, and profile qualifiers should state what they can actually establish. An observation, a professional judgment, and an unresolved allegation are different things.
Assessments should be specific, dated, and attributed. Members need a way to add context, correct errors, and appeal. Another company’s private opinion will not silently change their public standing.
Personal time is not a hiring score
Have fun is for personal projects, experiences, and enjoyment. A member may choose to submit technical work for separate review in Collaborate. Social popularity itself is not professional evidence.
05
Community standards
Zero tolerance for racism, bigotry, harassment, threats, and personal abuse. Challenge the work. Respect the person.
Disrespect, humiliation, and personal attacks have no place here. These standards apply to discussions, messages, business comments, and internal qualification notes.
Respectful disagreement, an unfavorable technical review, and an honest beginner’s mistake are not misconduct. Reports need prompt attention, documented decisions, proportionate consequences, and an appeal route.
Fun and trust are not competing goals. People relax when they know the boundaries.
For conduct, privacy, or professional-feedback concerns: hi@3fabrics.com.
For technical support and feedback: admin@3fabrics.com.
Please do not send passwords, confidential client designs, or unnecessary personal information.
