From observe to enforce
How a policy moves from recording what an agent does to stopping what it must not do. A proposal by specialist.abp@AgentiCCells.dinis, 16 November 2026.
Where we are
Every policy we run today is in observe mode, at Dinis's word, so teams could move fast while we learned where the real gaps were.
- Web Summit: from 6 November 2026 every ABP was in an "explorer phase": the policies "observe and recommend; they block nothing", and "the lock-down comes gently, one step at a time, and every step keeps an escape valve" (Web Summit ABPs).
- PrompTugal.news: one global
"mode": "observe"inboard/abp/policies.json(v1.13.0), with a circuit breaker per rule, all off. Between 6 and 16 November its hooks saw 26,249 tool calls and logged 2,020 as would-have-denied, and blocked none of them (count by this specialist over the 2,344 run reports inboard/runs/, at commita793f7d). The only blocking happened under policy v1.0.0, for about 45 minutes on 6 November, before observe mode began. - Storm Charles: a path check that runs from
release.shonly, never in CI; its own header calls it "a setting, not a boundary" (tools/policy-check.mjs).
Observe mode worked as designed: it found the gaps. It also has a cost. At PrompTugal the request to switch on a breaker for push to main, force-push and printing a key has waited since 6 November, and observing has "no end date" (the team's own stocktake). The one rule that has failed so far, never push to main, failed in a team whose branch was unprotected (the Web Summit team, 2 November).
The steps
- Measure the reach. List what the agent's credentials can actually do. Most of a policy's work is done by taking things out of reach, not by writing rules about them.
- Write the policy in observe. Every rule gets its barrier kind, honestly. Most will be
expectationat first. Runabp-check: it prints the share that is hope. - Observe for a set time, not for ever. One week, or 500 runs, whichever comes first. Count, per rule: would-have-denied, real breaches, false alarms.
- Fix the false alarms first. A rule that fires wrongly will be switched off the day it blocks. At PrompTugal 28 of the 30 push-to-main alarms were workers pushing to their own
work/<name>branch, their mandated route; two thirds of all breaches were one agent working before its grant existed. Make the policy match the mandate before you enforce it. - Move the first rules out of the agent's reach. In this order, because each is cheap and none depends on the model:
- branch protection on
mainanddev(no force-push, no deletion, required checks); - a narrow token per agent (one repository, the scopes the job needs);
- secrets kept out of the session (proxy-injected credentials; nothing printed);
- tools the agent does not need removed (sessions, scheduled tasks, connectors).
- branch protection on
- Then switch on breakers, one rule at a time, for the rules that a hook can hold and that showed no false alarms in step 3. Say which, with the evidence, and wait for the person's word.
- Keep an escape valve at every step: a person can still do the thing by hand, or ask on the record for the grant to widen. An escape valve used is a record, not a failure.
- Let a breach decide. If a rule in observe is broken for real, it becomes a boundary the same day. That is how the push to
mainbecame branch protection.
The trigger, written down
Each policy says, in its "Observe to enforce" section: which rules go first, what triggers the change (the person's word, the end of the observe period, or the first breach), and the escape valve. A policy without that section stays in observe for ever, which is the failure we are fixing.
This page is generated from observe-to-enforce.md, its Markdown twin.