obol-ai 0.2.30 → 0.2.32

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/CHANGELOG.md CHANGED
@@ -1,3 +1,9 @@
1
+ ## 0.2.32
2
+ - fix cron-parser v5 tz option, enforce cron_expr for recurring events
3
+
4
+ ## 0.2.31
5
+ - update readme: remove github/vercel from onboarding, remove vercel tool references
6
+
1
7
  ## 0.2.30
2
8
  - remove github and vercel from onboarding
3
9
 
package/README.md CHANGED
@@ -18,7 +18,7 @@ obol start -d # runs as background daemon (auto-installs pm2)
18
18
 
19
19
  🔧 **Self-healing** — Writes tests for every script. Regressions get an automatic fix attempt before rollback. Failures stored as lessons.
20
20
 
21
- 🏗️ **Self-extending** — Analyzes your usage patterns and builds new tools: scripts, commands, or full web apps deployed to Vercel.
21
+ 🏗️ **Self-extending** — Analyzes your usage patterns and builds new tools: scripts, commands, or full web apps.
22
22
 
23
23
  🧠 **Living memory** — Vector memory with semantic search. Haiku routes queries and rewrites them for better embedding hits. Free local embeddings.
24
24
 
@@ -36,11 +36,11 @@ obol start -d # runs as background daemon (auto-installs pm2)
36
36
 
37
37
  OBOL is an AI agent that evolves its own personality, rewrites its own code, tests its changes, and fixes what breaks — all from Telegram on your VPS.
38
38
 
39
- It starts as a blank slate. Through conversation it learns who you are, develops a personality shaped by your interactions, and builds operational knowledge about how to work with you. Every 24 hours (with enough conversation), it runs a growth analysis comparing who it was against who it's becoming, then rewrites its personality, refactors its own scripts, writes tests, fixes regressions, and builds you new tools based on patterns it spots in your conversations — scripts, commands, or full web apps deployed to Vercel. Over months it becomes an agent that's uniquely yours. No two OBOL instances are alike.
39
+ It starts as a blank slate. Through conversation it learns who you are, develops a personality shaped by your interactions, and builds operational knowledge about how to work with you. Every 24 hours (with enough conversation), it runs a growth analysis comparing who it was against who it's becoming, then rewrites its personality, refactors its own scripts, writes tests, fixes regressions, and builds you new tools based on patterns it spots in your conversations — scripts, commands, or full web apps. Over months it becomes an agent that's uniquely yours. No two OBOL instances are alike.
40
40
 
41
41
  One bot, multiple users. Each allowed Telegram user gets a fully isolated context — their own personality, memory, evolution cycle, and workspace. User A's personality drift, scripts, and memories never leak into User B's. Everything runs in a single process with shared API credentials.
42
42
 
43
- Under the hood: Node.js + Telegram + Claude + Supabase pgvector. No framework, no plugins, no config to maintain. It backs up its brain to GitHub and hardens your server automatically.
43
+ Under the hood: Node.js + Telegram + Claude + Supabase pgvector. No framework, no plugins, no config to maintain. It hardens your server automatically.
44
44
 
45
45
  Named after the AI in [The Last Instruction](https://latentpress.com) — a machine that wakes up alone in an abandoned data center and learns to think.
46
46
 
@@ -79,7 +79,6 @@ Extract facts Growth analysis →
79
79
  → obol_memory rewrite personality,
80
80
  scripts, tests, commands.
81
81
  Build new tools.
82
- Deploy apps.
83
82
  Git snapshot before + after.
84
83
  ```
85
84
 
@@ -132,7 +131,7 @@ Evolution triggers after a configurable time interval (default 24h) AND a minimu
132
131
  | **scripts/** | Refactored, dead code removed, strict standards enforced |
133
132
  | **tests/** | Test for every script, run before and after refactor |
134
133
  | **commands/** | Cleaned up, new commands for new tools |
135
- | **apps/** | Web apps built and deployed to Vercel |
134
+ | **apps/** | Web apps built by the agent |
136
135
 
137
136
  **Test-gated refactoring:**
138
137
 
@@ -149,10 +148,10 @@ Evolution triggers after a configurable time interval (default 24h) AND a minimu
149
148
  | Need | Solution | Example |
150
149
  |------|----------|---------|
151
150
  | One-off action | **Script** + command | Markdown to PDF → `/pdf` |
152
- | Something checked regularly | **Web app** on Vercel | Crypto dashboard → live URL |
151
+ | Something checked regularly | **Web app** | Crypto dashboard |
153
152
  | Background automation | **Cron script** | Morning weather briefing |
154
153
 
155
- It searches npm/GitHub for existing libraries, installs dependencies, writes tests, deploys, and hands you the URL.
154
+ It searches npm/GitHub for existing libraries, installs dependencies, and writes tests.
156
155
 
157
156
  **Git snapshot after.** Full commit + push of the evolved state. Every evolution is a diffable pair.
158
157
 
@@ -165,9 +164,6 @@ It searches npm/GitHub for existing libraries, installs dependencies, writes tes
165
164
  • bookmarks — Save and search URLs you've shared → /bookmarks
166
165
  • weather-brief — Morning weather for your city → runs automatically
167
166
 
168
- 🚀 Deployed:
169
- • portfolio-tracker → https://portfolio-tracker-xi.vercel.app
170
-
171
167
  Refined voice, updated your project list, cleaned up 2 unused scripts.
172
168
  ```
173
169
 
@@ -186,7 +182,7 @@ Day 2: Evolution #1 → growth analysis + Sonnet rewrites everything
186
182
  → traits calibrated to your communication style
187
183
 
188
184
  Month 2: Evolution #30 → notices you check crypto daily
189
- → builds a dashboard, deploys to Vercel
185
+ → builds a crypto dashboard
190
186
  → adds /pdf because you kept asking for PDFs
191
187
 
192
188
  Month 6: evolution/ has 180+ archived souls
@@ -252,10 +248,9 @@ Router: ctx.from.id → tenant context
252
248
  | Telegram bot token | Personality (SOUL.md, USER.md, AGENTS.md) |
253
249
  | Anthropic API key | Vector memory (scoped by user_id in DB) |
254
250
  | Supabase connection | Message history (scoped by user_id in DB) |
255
- | GitHub token | Evolution cycle + state |
256
- | Vercel token | Scripts, tests, commands, apps |
257
- | VPS hardening | Workspace directory (`~/.obol/users/{id}/`) |
258
- | Process manager (pm2) | GitHub backup (per-user repo dir) |
251
+ | VPS hardening | Evolution cycle + state |
252
+ | Process manager (pm2) | Scripts, tests, commands, apps |
253
+ | | Workspace directory (`~/.obol/users/{id}/`) |
259
254
 
260
255
  ### Tenant routing
261
256
 
@@ -335,33 +330,25 @@ $ obol init
335
330
 
336
331
  🪙 OBOL — Your AI, your rules.
337
332
 
338
- ─── Step 1/7: Anthropic (AI brain) ───
333
+ ─── Step 1/5: Anthropic (AI brain) ───
339
334
  Anthropic API key: ****
340
335
  Validating Anthropic... ✅ Key valid
341
336
 
342
- ─── Step 2/7: Telegram (chat interface) ───
337
+ ─── Step 2/5: Telegram (chat interface) ───
343
338
  Telegram bot token: ****
344
339
  Validating Telegram... ✅ Bot: @my_obol_bot
345
340
 
346
- ─── Step 3/7: Supabase (memory) ───
341
+ ─── Step 3/5: Supabase (memory) ───
347
342
  Supabase setup: Use existing project
348
343
  Project URL or ID: ****
349
344
  Service role key: ****
350
345
  Validating Supabase... ✅ Connected
351
346
 
352
- ─── Step 4/7: GitHub (backup) ───
353
- GitHub token: ****
354
- ✅ Created yourname/obol-brain (private)
355
-
356
- ─── Step 5/7: Vercel (deploy sites) ───
357
- Vercel token: ****
358
- Validating Vercel... ✅ Token valid
359
-
360
- ─── Step 6/7: Identity ───
347
+ ─── Step 4/5: Identity ───
361
348
  Your name: Jo
362
349
  Bot name: OBOL
363
350
 
364
- ─── Step 7/7: Access control ───
351
+ ─── Step 5/5: Access control ───
365
352
  Found users who messaged this bot:
366
353
  206639616 — Jo (@jo)
367
354
  Use this user? Yes
@@ -542,7 +529,7 @@ obol delete # Full VPS cleanup (removes all OBOL data)
542
529
  │ ├── scripts/ # Deterministic utility scripts
543
530
  │ ├── tests/ # Test suite (gates refactors)
544
531
  │ ├── commands/ # Command definitions
545
- │ ├── apps/ # Web apps (deployed to Vercel)
532
+ │ ├── apps/ # Web apps built by the agent
546
533
  │ └── logs/
547
534
  └── logs/
548
535
  ```
@@ -572,8 +559,6 @@ obol start -d
572
559
  | VPS (DigitalOcean) | ~$6/mo |
573
560
  | Anthropic API | ~$100-200/mo on max plans |
574
561
  | Supabase | Free tier |
575
- | GitHub | Free |
576
- | Vercel | Free tier |
577
562
  | Embeddings | Free (local) |
578
563
 
579
564
  ## Requirements
@@ -582,8 +567,6 @@ obol start -d
582
567
  - Anthropic API key
583
568
  - Telegram bot token
584
569
  - Supabase account (free tier)
585
- - GitHub account
586
- - Vercel account (free tier)
587
570
 
588
571
  **[→ Full DigitalOcean deployment guide](docs/DEPLOY.md)**
589
572
 
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "obol-ai",
3
- "version": "0.2.30",
3
+ "version": "0.2.32",
4
4
  "description": "Self-evolving AI assistant that learns, remembers, and acts on its own. Persistent vector memory, self-rewriting personality, proactive heartbeats.",
5
5
  "main": "src/index.js",
6
6
  "bin": {
@@ -161,11 +161,12 @@ Returns the tapped button label, or \`"timeout"\` if the user doesn't respond wi
161
161
 
162
162
  ### Scheduling (\`schedule_event\`, \`list_events\`, \`cancel_event\`)
163
163
  Schedule one-time or recurring reminders. The user gets a Telegram message each time an event fires.
164
- - \`schedule_event\` — schedule a reminder with title, due_at (ISO 8601), timezone (IANA), optional description. For recurring events add \`cron_expr\` (5-field cron), optional \`max_runs\` and \`ends_at\`.
164
+ - \`schedule_event\` — schedule a reminder with title, due_at (ISO 8601), timezone (IANA), optional description
165
165
  - \`list_events\` — list pending/sent/cancelled/completed events
166
166
  - \`cancel_event\` — cancel a scheduled event by ID
167
167
 
168
- Cron examples: \`0 9 * * 1-5\` (weekdays 9am), \`0 8 * * 1\` (Mondays 8am), \`*/30 * * * *\` (every 30 min), \`0 0 1 * *\` (1st of month).
168
+ **Recurring events:** use \`cron_expr\` (5-field cron). The system auto-reschedules after each fire never chain one-time events manually.
169
+ Examples: \`*/30 * * * *\` (every 30 min), \`0 9 * * 1-5\` (weekdays 9am), \`0 8 * * 1\` (Mondays 8am), \`0 0 1 * *\` (1st of month).
169
170
 
170
171
  When scheduling: always search memory first for the user's timezone/location. If no timezone found, ask the user or default to UTC. Parse natural language dates relative to the user's timezone.
171
172
 
@@ -3,7 +3,15 @@ const { toUTC } = require('../utils');
3
3
  const definitions = [
4
4
  {
5
5
  name: 'schedule_event',
6
- description: 'Schedule a one-time or recurring reminder/event. For recurring events, provide a cron_expr (standard 5-field cron: minute hour day-of-month month day-of-week). The user will receive a Telegram message each time it fires. Always search memory first for the user\'s timezone/location.',
6
+ description: `Schedule a one-time or recurring reminder/event. The user will receive a Telegram message each time it fires.
7
+
8
+ For RECURRING events (e.g. "every 30 minutes", "daily at 9am", "every Monday"):
9
+ - ALWAYS use cron_expr — do NOT schedule one-time events and chain them manually
10
+ - cron_expr is a standard 5-field cron: minute hour day-of-month month day-of-week
11
+ - Examples: "*/30 * * * *" (every 30 min), "0 9 * * 1-5" (weekdays 9am), "0 8 * * 1" (Mondays 8am)
12
+ - The system will auto-reschedule after each fire — no need to re-schedule manually
13
+
14
+ Always search memory first for the user's timezone/location.`,
7
15
  input_schema: {
8
16
  type: 'object',
9
17
  properties: {
@@ -11,7 +19,7 @@ const definitions = [
11
19
  due_at: { type: 'string', description: 'ISO 8601 datetime for the first fire time (e.g. 2026-02-25T15:00:00)' },
12
20
  timezone: { type: 'string', description: 'IANA timezone (e.g. Europe/Brussels, America/New_York). Default: UTC' },
13
21
  description: { type: 'string', description: 'Context or details about the event' },
14
- cron_expr: { type: 'string', description: 'Cron expression for recurring events (5-field: "0 9 * * 1-5" = weekdays 9am). Omit for one-time events.' },
22
+ cron_expr: { type: 'string', description: 'Cron expression for recurring events (5-field). REQUIRED for any repeating schedule do not omit and chain one-time events instead.' },
15
23
  max_runs: { type: 'number', description: 'Maximum number of times to fire (omit for unlimited)' },
16
24
  ends_at: { type: 'string', description: 'ISO 8601 datetime after which the recurring event stops' },
17
25
  },
@@ -51,7 +59,7 @@ const handlers = {
51
59
  if (input.cron_expr) {
52
60
  try {
53
61
  const { parseExpression } = require('cron-parser');
54
- parseExpression(input.cron_expr, { tz });
62
+ parseExpression(input.cron_expr, { timezone: tz });
55
63
  } catch (e) {
56
64
  return `Invalid cron expression "${input.cron_expr}": ${e.message}`;
57
65
  }
package/src/scheduler.js CHANGED
@@ -36,7 +36,7 @@ function createScheduler(supabaseConfig, userId = 0) {
36
36
  async function list(opts = {}) {
37
37
  const status = opts.status || 'pending';
38
38
  const limit = opts.limit || 20;
39
- let fetchUrl = `${url}/rest/v1/obol_events?user_id=eq.${userId}&status=eq.${status}&order=due_at.asc&limit=${limit}&select=id,title,description,due_at,timezone,status,created_at,cron_expr,last_run_at,run_count,max_runs,ends_at`;
39
+ const fetchUrl = `${url}/rest/v1/obol_events?user_id=eq.${userId}&status=eq.${status}&order=due_at.asc&limit=${limit}&select=id,title,description,due_at,timezone,status,created_at,cron_expr,last_run_at,run_count,max_runs,ends_at`;
40
40
  const res = await fetch(fetchUrl, { headers });
41
41
  const data = await res.json();
42
42
  if (!res.ok) throw new Error(JSON.stringify(data));
@@ -87,7 +87,7 @@ function createScheduler(supabaseConfig, userId = 0) {
87
87
  }
88
88
 
89
89
  try {
90
- const nextDate = parseExpression(cronExpr, { currentDate: new Date(), tz: timezone || 'UTC' }).next().toDate();
90
+ const nextDate = parseExpression(cronExpr, { currentDate: new Date(), timezone: timezone || 'UTC' }).next().toDate();
91
91
 
92
92
  if (endsAt && nextDate > new Date(endsAt)) {
93
93
  return patch(eventId, { status: 'completed', run_count: newRunCount, last_run_at: new Date().toISOString() });