Repository navigation
StudySprint v0.2.0
Pre-release
Pre-release
[0.2.0] — 2026-10-31
Second versioned state: MINOR feature batch implementing all five
FEATURES.md Now items, automated re-sync parity verification, and
refreshed evidence snapshots.
Provider credit: Jules, PR #2635, session 17326020844735872479.
Added
- Trust receipt formatting (
format_trust_receiptinstudy_tools.py):
surfaces verified course metadata, deadline counts, source path, and
short content hash before student confirmation; wired into agent receipt
trust_receiptpayload. - Today view card (
render_today_viewinstudy_tools.py): renders
single-screen card of today's study sessions with estimated minutes,
"Start Today's Sprint" primary action, and guilt-free "I missed this"
action; wired into agent receipttoday_view. - Plan-budget reflow (
reflow_plan_budgetinstudy_tools.py):
dynamically re-scales session duration (e.g. 25 min vs 45 min) to match
the student's real daily schedule. - Daily check-in data model (
record_daily_checkininstudy_tools.py):
tracks session completion (done/partial/missed), re-plan recommendations,
and streaks framed as active/recovered weeks rather than broken chains. - Plain-language mode & accessible statuses (
plain_language_action,
format_accessible_status,PRIVACY_STATEMENTinstudy_tools.py):
plain-language explanations without jargon, high-contrast accessible
gate decision tags ([✓ ALLOW],[⚠ ABSTAIN],[⛔ ESCALATE]), and
local-device privacy guarantee in every receipt. - Automated re-sync parity auditor (
tools/resync_parity.pyand
tests/test_resync_parity.py): computes Git blob object SHAs
(sha1("blob " + len + "\0" + content)), comparing monorepo tree
against upstreamtimerloggedout-spec/studysprint(GitHub API or local
dir) with offline baseline manifest support. - Extended Strands agent tool declarations in
agent.py: added tools
get_today_view,adjust_time_budget, andcheckin_session.
Changed
- Receipts: schema
studysprint-agent-receipt/v1updated to carry
agent_version: "0.2.0",decision_accessible,
plain_language_explanation,trust_receipt,today_view, and
privacy. - VERSION bumped to 0.2.0.
Evidence
- Tests: 17 passed (offline suite:
tests/test_studysprint.py+
tests/test_resync_parity.py). - Demo decisions (
evidence/demo-run.jsonatagent_version: "0.2.0"):
scenario 1 ABSTAIN on asserted-only evidence; scenario 2
ALLOW_RECOMMENDATION on fetched syllabus (12 sessions, week-1 quiz,
re-plan, trust receipt, today view); scenario 3 ESCALATE on
credential_mutation.