source pages
the deeper bibliography: egor, levin, miessler, and the source-system overview.
this page is the live scan-layer for the project: what looks settled, what is still in motion, and what deserves attention now. it is not the canonical doctrine page and not the source bibliography. those live in separate surfaces.
the pages should not compete with each other. each one should solve a different reading job.
the stable canonical layer: mission, wedge, v1, principles, trust ladder, business posture, and non-goals.
evidence-backed strategic claims, confidence, caveats, falsifiers, and roadmap implications.
the process layer for how the constitution gets updated, what is done, and what is still missing.
how truth moves: update rhythms, source hierarchy, kill conditions, and delta-only review loops.
the working shortlist of external harnesses worth trying, studying, or skipping for secondme.
the deeper bibliography: egor, levin, miessler, and the source-system overview.
the main fix for duplication is to keep each page narrow.
mission, wedge, v1, principles, trust ladder, business posture, non-goals.
current posture, strongest beliefs, open tensions, and immediate focus.
claim, evidence, caveat, falsifier, implication, confidence.
source hierarchy, review rhythm, open-loop discipline, and promotion rules.
agent roles, card schemas, section thresholds, current status, and next implementation steps.
the point of the overview is not to repeat all the doctrine. it is to keep the active field visible: what seems settled, what is still soft, and what needs disciplined pressure next.
the system is valuable when it coordinates reality: priorities, people, approvals, timing, and follow-through.
memory, orchestration, and decision surfaces appear more leverage-bearing than model chasing for real workflows.
the strongest axis still looks like complexity x stakes x agency, with HNW as one likely first segment.
the product still needs one sharp flow where a user grants read-only access to live context, gets immediate insight, and sees tailored leverage already in motion.
this is the information architecture shift suggested by the team discussion: less polished certainty, more readable tension.
every active bet should show confidence, test, owner, review date, and the condition under which the team would kill it.
surface the strongest confirming signal, strongest disconfirming signal, and the weird new signal that changed the map.
doctrine should be brief. the evidence, caveats, and falsifiers should live in a linked proof layer.
track what changed outside in the last 30 days so the team does not keep solving a dead problem with beautiful language.
every expensive hanging loop needs an owner, next step, and review date. loops without owners decay into mood.
a pre-pm-fit dashboard should be clear, alive, and slightly uncomfortable. otherwise it becomes a status object.
one click for strategic belief. one click for team operating discipline. then deeper source pages for provenance.
the stable doctrine page: what secondme is, who it is for, what v1 is, and which principles and non-goals govern decisions.
the evidence layer: claims, caveats, falsifiers, and roadmap implications behind the strongest beliefs.