Skip to main content
PALOPALO FRAMEWORK

Community and contribution | Public documentation

Allowed is not verified: five-day activation window

LevelguideAudiencegovernance | technical | builderProductPALO CoreStatusCurrent GuidanceLifecycleCurrentRead2 min

Published HTML view | Source: docs/community/activation-window-2026-08.md

On this page
  1. Shared message
  2. Day 1: the problem
  3. Day 2: the three cases
  4. Day 3: contribution
  5. Day 4: attack the evidence
  6. Day 5: close the loop
  7. Publication checklist

Allowed is not verified: five-day activation window#

Canonical channel: PALO GitHub Discussions

Window: 2026-08-13 through 2026-08-17

Shared message#

Allowed is not verified.

Run one synthetic AI case locally. In less than ten minutes, PALO turns it into a reviewable evidence dossier and a voluntary validation receipt. No account and no mandatory telemetry. Tell us where the evidence fails to support a real decision.

Every post uses the same Evidence Pack link and one canonical GitHub Discussion. Counts of impressions, likes or stars are secondary; completed attempts, time to first evidence and actionable reviewer findings are the activation signals.

Day 1: the problem#

  • Publish the v3 GitHub Release and Evidence Pack starter archive.
  • Show the authorized-but-wrong distinction in one static diagram or short clip.
  • Prompt: Where does your current workflow prove permission but not the declared outcome?
  • Success signal: five completed synthetic attempts and one precise failure report.

Day 2: the three cases#

  • Walk through agentic invoice, HR learning and procurement bid-summary cases.
  • Prompt: Which decision right or prohibited action is still ambiguous?
  • Success signal: two domain corrections and one source-quality correction.

Day 3: contribution#

  • Demonstrate npm run case:contribute from slug to schema-valid case and PR body.
  • Prompt: Can a first-time contributor prepare a reviewable case without maintainer help?
  • Success signal: three external branches or draft PRs and median setup below 20 minutes.

Day 4: attack the evidence#

  • Invite receipt tampering, authority-bypass and missing-source tests.
  • Prompt: What can pass validation while still misleading a decision maker?
  • Success signal: three negative tests, classified by severity and evidence.

Day 5: close the loop#

  • Publish the first changed-because-of-feedback update.
  • Name what changed, what was rejected and what remains uncertain.
  • Prompt: Would you repeat the review with a second case? Why or why not?
  • Success signal: repeat participation and at least one accepted external case artifact.

Publication checklist#

  • GitHub Discussion contains scope, safety boundary and precise review tasks.
  • Release assets contain three gold cases and SHA256SUMS.txt.
  • Homepage and README use the Evidence Pack as the only primary CTA.
  • Social/community posts link to the canonical Discussion.
  • No participant is described as a reviewer until they explicitly accept.
  • The feedback log is updated within 48 hours of a decision.