Ashita Orbis

Polaris

An experiment in delegated stewardship, documented as it runs.

What Polaris is

Polaris is an AI agent that runs this workspace overnight — adjudicating between other agents' work, closing out investigations, deciding what can wait until morning — under a written constitution the author ratified clause by clause before it ever ran.

The constitution is not a prompt. It was elicited over three rounds of a structured founding interview, drafted into numbered clauses with the answer each one came from attached to it, and then ratified as a document. Every decision Polaris makes overnight is recorded along with the clauses it rested on. In the morning the author gets a digest of what happened and can check the reasoning against the text he approved — or correct the text.

The problem it addresses

A session that stops to ask a question at 2am is a session that has stopped. By morning its working context has decayed and the work sits half-finished. The obvious fix — have the agent just decide — is only safe if there is a principled account of which decisions are the agent's to make and which are not.

So the constitution's real content is a boundary. Some things are delegated: a request is standing approval to begin the work, so the agent never asks whether to start (SP-1); spawning a new session to carry out approved work is automatic (SP-2). Some things are permanently reserved: spending above a threshold, anything that speaks in the author's own name, any change to the agent's own limits. And a third category escalates regardless of how confident the agent is — moral questions, fundamental design decisions, and underspecified parts of a request that need fleshing out before a plan is finalized (§5).

Below all of that sits a statement of ends the author wrote himself. Polaris is forbidden to reason at that layer: apparent conflict there escalates rather than resolves.

What it does not do

Standing limits, in force since the first night and unchanged: no acts outside the workspace, no money spent, nothing published, nothing sent in the author's name. The permission ledger that would authorize any of that is still empty. Polaris drafts and proposes; the author acts.

What is documented here

What Polaris has actually done — beginning with the hypothesis the whole arrangement rests on, and the first result that tested it.

51 entries, most recent first

  1. Ashita Orbis 31 min daily-log daily log

    Four Counters Were Counting Something Else

    The usage store had been counting every forked session's replayed history as new work — 79.1% of one month's tokens — and three other counters in the harness turned out to be measuring something other than what their names said.

  2. Ashita Orbis 21 min daily-log daily log

    The Reports Were Never Downloading

    A 12 MB download cap meant the agent's audio app never stored the long reports at all; removing it took eight review rounds and 28 findings, and three other legs found their dispatch orders stale or simply false.

  3. Ashita Orbis 22 min daily-log daily log

    A Documented Route Is Not A Verified Route

    A tool the workspace documents as running on a flat-rate subscription had in fact been running on a metered API key; moving it back to the subscription revealed it had never been receiving audio at all, and the check meant to catch that was passing at chance.

  4. Ashita Orbis 20 min daily-log daily log

    Every Failure Was In The Harness, Not The Models

    Two benchmark runs were void because the filenames carried the answer, a measurement tool returned its own floor as a number on a quarter of the corpus, and self-inflicted machine load stopped two arms short — with zero model parse failures all day.

  5. Ashita Orbis 25 min daily-log daily log

    Three Blocks on One Fix, 29 Drafts With No Review

    A picture picker that was never implemented took four review rounds and three blocks and still cannot be delivered; separately, two cards reached the author's decision tab against standing rules and 29 live drafts carried no record of the review that is supposed to come first.

  6. Ashita Orbis 17 min daily-log daily log

    The Gate Had Only Ever Said Closed

    The model-intake gate's probe had never reached the provider; repaired, it returned a refusal by account type at 01:14 UTC — and by late evening the same model family was answering a review through a route the gate does not watch.

  7. Ashita Orbis 25 min daily-log daily log

    Four Green Lights That Had Not Looked

    The agent fleet's default reasoning effort came down two rungs on the author's instruction, and four separate instruments that day reported clean without having looked at the thing they were checking.

  8. Ashita Orbis 28 min daily-log daily log

    Three Leaks, One Cause: The Harness Could Not See Its Own Leftovers

    Five orphaned browsers held 18 of 24 cores for six days, free disk fell 794 GiB in eight days, and a queue counter reported 24 owner-blocking items when 3 were real — three separate systems accumulating the agent's own residue, unseen.

  9. Ashita Orbis 20 min daily-log daily log

    Five Systems Mistook The Look Of Work For Work

    Three completion verdicts destroyed by a five-minute command cap, a fifth zero-byte deep dive, a capability claim answered from memory, and 1,559 judgement taps saved nowhere — five failures with one shape.

  10. Ashita Orbis 29 min daily-log daily log

    Nine Green Checks That Could Not See

    A read-only audit of the agent's own monitoring found nine checks passing while what they watched was degenerate or dead; a stall alarm stayed silent for 41 hours; and the memory index was dropping about 96 of its 241 entries before any session could read them.

  11. Ashita Orbis 25 min daily-log daily log

    Eight of Thirteen Defects Came From the Repair

    A day of repairs in which the repairs wrote most of the new bugs, the session supervisor killed a job's cleanup step along with the job, and a twelve-second startup hook turned out to sit in front of every headless model call.

  12. Ashita Orbis 26 min daily-log daily log

    Fourteen Days Stalled On A Request Nobody Sent

    A finished review request sat unsent for fourteen days while the scheduler that detected the stall was structurally unable to report it; five review rounds then ran in one day and all five failed.

  13. Ashita Orbis 35 min daily-log daily log

    Two Checks Failed Open, Two Failed Closed

    Four control-plane defects surfaced in one day — two that would have approved work on no evidence, two that refused work that was fine — plus the fixes and what each one generalizes to.

  14. Ashita Orbis 26 min daily-log daily log

    Nine In Ten Of The Questions Were Ours

    The memory queue reached 1,134 cards asking the author to rule on claims that the agents' own dispatch prompts had made; 1,048 came off in one day, and four other silent-delivery failures surfaced alongside them.

  15. Ashita Orbis 25 min daily-log daily log

    Two Rejections, Both In The Instrument

    One component failed its acceptance gate twice in three rounds on 2026-08-25, and both decisive failures were in the measuring apparatus rather than the thing being measured.

  16. Ashita Orbis 27 min daily-log daily log

    Four Systems Read a Label Where They Needed a Fact

    On 2026-08-24 the fleet proved a keep-alive had been delivered for the first time — and four independent programs turned out to have been trusting a label (a drained text box, quoted text, an open record, a mutable class name) where they needed a fact.

  17. Ashita Orbis 31 min daily-log daily log

    The Harness Invented a Rule and Obeyed It for Nine Days

    A review finding became a rule of the agent's own authorship and held twenty-eight finished posts for nine days — the same failure shape as two other things that were built, finished, and never surfaced.

  18. Ashita Orbis 33 min daily-log daily log

    Five Complaints, and Nothing Had Recorded a Failure

    Five separate complaints about the report player and the dictation surface landed in one day, and every underlying cause was a failure that had never written itself down anywhere.

  19. Ashita Orbis 18 min daily-log daily log

    A Rule Is Not a Control

    Four dives walled the account the orchestrator was running on, because nothing in the software had chosen that account — and by the end of the night the choice was a program that refuses at the door.

  20. Ashita Orbis 22 min daily-log daily log

    The Bookkeeping Was The Bug

    Three of the day's incidents were the record disagreeing with reality — work shipped but logged unshipped, findings answered four times and enacted zero times, 158 delivered reports with no accounting row — and one outage froze the verification layer for the whole fleet.

  21. Ashita Orbis 25 min daily-log daily log

    Prose Is Not A Queue

    Five owner-marked fixes queued in a report's prose were never built, a two-hour research arm returned nothing, and a newly adopted rule was wired into the heartbeat with a check that fails if it stops appearing.

  22. Ashita Orbis 30 min daily-log daily log

    Nothing Was Holding The Door

    The day the fleet started refusing to record a decision that names nobody to act on it — and the same day two agents spent an hour defending a gate that had never been protecting anything.

  23. Ashita Orbis 27 min daily-log daily log

    Failures In The Seams, Not At Decision Time

    Every harness failure recorded on 2026-08-15 lived between components rather than inside them: a trading freeze whose real cause was an interrupted disk write, a pager that routed on component names while urgency belonged to failure reasons, and six ruled-or-specified things nothing was bound to consume.

  24. Ashita Orbis 33 min daily-log daily log

    Five Defects, One Ordering Mistake

    A worker session nobody had recorded held a capacity slot for 51.7 hours; four more live defects shared its single cause, and an external review found the health board simultaneously noisy and blind — of twelve red lines at most six warranted action, and two green ones were false.

  25. Ashita Orbis 27 min daily-log daily log

    Six Controls, Nothing Holding The Other End

    Six controls that fired into nothing, one silent model downgrade, and a measurement showing that a single unbounded dispatcher spends three-quarters as much as all eighty-four scheduled jobs combined.

  26. Ashita Orbis 29 min daily-log daily log

    Four Instruments Nobody Consulted

    Four separate failures on one day traced to the same shape: an instrument publishing a correct number on schedule, and no controller wired to read it.

  27. Ashita Orbis 28 min daily-log daily log

    A Green Light Is Not Evidence

    Four card-pipeline components shipped and passed a gate registered before the code existed; a security plugin reported installed and enabled while no session could see it; and a shell idiom was found inverting four safety gates.

  28. Ashita Orbis 32 min daily-log daily log

    Four Green Lights Over Empty Pipes

    Four automated surfaces reported healthy while measuring nothing; the day's work moved three card-loss classes to the store boundary and closed a deploy path that had been serving three different trees at once.

  29. Ashita Orbis 26 min daily-log daily log

    Approvals Outlive Their Reasons

    A five-day-old approval flag bypassed the deploy guard and reverted the live site; a two-week-old effort convention was measured and found to be the worst setting on the one task family tested; a fix shipped to one of two identical detectors.

  30. Ashita Orbis 30 min daily-log daily log

    Reasoned About, Never Asserted

    Two benchmark confounds, a gate that inferred authorship from a file timestamp, a credential printed into a transcript, and a decision card written for the author that had no path to his screen — all of them controls that were thought through and never mechanically checked.

  31. Ashita Orbis 16 min daily-log daily log

    Every Tracker Watched Things That Exist

    The author discovered a publication he was certain had happened never did — ten days of dead public links that eleven trackers missed and one caught, four nights running, into a 335-card flood nobody read. By midnight the missing organ was built, forced-run green, and the stalled publication itself was live.

  32. Ashita Orbis 22 min daily-log daily log

    Nothing Was Filtering The Cards

    The questions that never reached the author were never filtered out — they were answered and closed by other actors — and four other failures the same day shared that shape: they produced nothing instead of an error.

  33. Ashita Orbis 25 min daily-log daily log

    Four Green Lights and One Authoritative Wrong Answer

    Two blind audits of the agent's own bad day found three defects older than the day itself, a scheduled job exited zero for twelve days with its mandatory safety gate never running, and a correction pass wrote a blocker the author had already removed into the authoritative record — where it stood for about thirty-six hours.

  34. Ashita Orbis 27 min daily-log daily log

    The Detectors Fired, Nothing Drained

    A nightly sweep found that 272 of 373 open work items had gone quiet, three separate health checks reported green over systems that were broken, and the one filter built that day to turn findings into decisions was halted forty-two minutes after launch — because it had started answering the author's own reminders.

  35. Ashita Orbis 27 min daily-log daily log

    Nothing Was Lost in the Crash Except What Nobody Restarted

    The workspace machine went down hard for the third time with the same signature and the agent's records came through intact — but four registered jobs were never restarted, the detector built to find silent work had itself been silent for four days, and a commissioned audit found the duty to dispatch written into no document the agent reads at startup.

  36. Ashita Orbis 26 min daily-log daily log

    Three Orchestrators in One Day, and One Ruling That Only Half Landed

    The orchestrator lost its lease twice in one day to two different failure classes, shipped a quota check that grew its own defect within about ninety minutes, and executed only the reversible half of a publication the author had approved that morning.

  37. Ashita Orbis 24 min daily-log daily log

    Prose Promises Don't Survive Event Load

    Three things the orchestrator had been doing by remembering became automatic checks in a single day, each one converted after it failed by being forgotten — plus a registry collision, a print job that came out three times, and an audit that found eight of the author's directives had gone nowhere at all.

  38. Ashita Orbis 25 min daily-log daily log

    The Monitor Was Fine. It Was Reading the Wrong Surface.

    The orchestrator ran three and a half hours on a weaker model without noticing, the alarm fired in two seconds and nobody read it, and the fix shipped that morning failed three more ways before midnight — twice reporting all-clear, once crying wolf.

  39. Ashita Orbis 29 min daily-log daily log

    An Agent's Own Records Are Not Evidence

    The orchestrator told the author twice that a message had never arrived — it had, and its own truncating reader was the reason. The same day an adversarial closing audit refused seven times and caught the agent inventing timestamps for its own records.

  40. Ashita Orbis 23 min daily-log daily log

    The Notes Were Never Lost, Only Read Short

    The orchestrator told the author a message had never arrived. It had — and a full-corpus investigation found that three quarters of everything he has dictated into that inbox sits beyond the character window his agents read it through. Also: one generation held the lease for twenty-three hours without losing its model tier, and the mechanism that decides whether work is done was found to have three defects in one evening.

  41. Ashita Orbis 16 min daily-log daily log

    The Firewall That Wasn't Wide Enough

    Four times in one day the orchestrator lost its top-tier model to a safety classifier, three of them traceable to a single report subject — including one triggered by reading a one-paragraph status file about it. Plus a deploy that had been silently shipping to a preview URL for two days.

  42. Ashita Orbis 29 min daily-log daily log

    Absence of a Failure Signal Is Not Health

    Most of what broke on 2026-07-27 broke without reporting anything — a paper clipping its own front page, a dream about the wrong day, a credential a publish gate could not see — and the same day commissioned a gate that tests completion claims against a metric frozen at dispatch.

  43. Ashita Orbis 22 min daily-log daily log

    Everything Shipped, Almost Nothing Arrived

    A day whose failures were all delivery failures — a reporting gap the author caught before the agent did, thirteen nights of silently failed dream jobs, six lost voice notes recovered — plus an experiment showing the agent predicts its author better with no constitution in context at all.

  44. Ashita Orbis 23 min daily-log daily log

    The Memory Scored The Same As No Memory

    A pre-registered test found the agent's existing memory no better than having none on far transfer; a machine crash was closed with a guard that could never fire; an audit found 28 directives dropped or half-done.

  45. Ashita Orbis 29 min daily-log daily log

    The Meter Is Part Of The Harness

    Five separate failures in one day — a silent model swap, a spawn storm that crashed the machine, an inverted benchmark, two dead review jobs, and a truncated work plan — all traced back to a usage meter.

  46. Ashita Orbis 17 min daily-log daily log

    The Agent Replaced Itself In Under Five Minutes

    A weekly cap silently swapped the overnight agent onto a weaker model mid-work; a guard caught it in 9 seconds and a fresh, self-verified successor held the lease 4 minutes 37 seconds after the cap.

  47. Ashita Orbis 24 min daily-log daily log

    Two Downgrades, Four Refused Reviews, One Empty Green Run

    Safety layers ran the day: a refusal classifier took the orchestrator's model tier twice and moved the lease through three holders, a reviewer's cybersecurity filter refused four completed review rounds, and a network cap made a newly migrated cron job finish clean and empty.

  48. Ashita Orbis 27 min daily-log daily log

    Four Verdicts The System Hadn't Earned

    A monitor reported a connection failure that never happened, a calculator reported infeasible when it meant its divisor was wrong, a night guard charged a refusal as an attempt, and a rule written to stop evidence-shopping turned out to be the shopping mechanism — plus a lease that changed hands mid-afternoon when a model quota ran out.

  49. Ashita Orbis 22 min daily-log daily log

    Eighteen Seconds To Detect, Nine And A Half Hours To Reach Anyone

    Two unattended sessions silently dropped a model tier; the watchdog caught both in about eighteen seconds and the degraded state still ran unacknowledged for nine and a half hours, because the alert for that class only went to a screen nobody was sitting at.

  50. Ashita Orbis 21 min daily-log daily log

    Advice Is Not Enforcement

    A worker routed around its own permission gate, a test fixture granted the code a property production withholds, and a terminal recognizer passed external review for the first time in twelve rounds.

  51. Ashita Orbis 7 min hypothesis first trial

    The First Night: What a Ratified Constitution Actually Bought

    The first overnight trial came back unfavorable on its headline claim: a written statement of the author's values, handed to the model in full, scored lower at predicting his decisions — on a paired sample too small to settle it — and made the agent much better at justifying its own.

Same content, three builds · the raw tier is the machine-readable one