Andy Vault Home

KPI Snapshot

• MCP: 🟢 `obsidian.andotherstori.my.id/mcp`
• Daily note cron: 🟢 06:00
• Meridian daily review cron: 🟢 06:05
• Insight sync cron: 🟢 06:10
• Vault path: `/home/ubuntu/vaults/andy`

Core Dashboards

• [[Projects/Meridian/Project Dashboard]]
• [[Projects/VPS/Project Dashboard]]
• [[Projects/Hermes/Project Dashboard]]

Reviews

• [[Projects/Meridian/Daily Reviews/2026/07/2026-07-08]]
• [[Projects/Meridian/Weekly Reviews]]

Daily Workflow

• [[Daily/Today]]
• [[Templates/Daily Note Template]]
• [[Templates/Meridian Daily Review Template]]
• [[Templates/Meridian Weekly Review Template]]
• [[Inbox/Quick Capture]]

Operations

• [[Areas/Operations/Meridian]]
• [[Areas/Operations/VPS]]
• [[Areas/Operations/Hermes]]
• [[Areas/Operations/VPS/Incident Log]]

Knowledge / Imports

• [[Inbox/Session Knowledge Import]]
• [[Inbox/Hermes Daily Insights]]
• [[Resources/Runbooks/Obsidian MCP Setup]]
• [[Resources/Runbooks/VPS Recovery Incident]]

Today

Use template: [[Templates/Daily Note Template]]

Top 3

Notes

Hermes Daily Insights

Updated: 2026-07-08 04:38:00Z

Highlights

OBSIDIAN

• [assistant] Saya lanjut bikin **otomatis**, bukan cuma file statis: 1. daily note auto-create tiap pagi 2. sync insight penting dari chat Hermes ke Obsidian tiap pagi Dashboard/detail page + runbook sudah jadi tadi. Sekarang saya pasang script + cron.
• [user] 1. bikin note harian otomatis tiap pagi 2. sync insight penting dari chat Hermes ke Obsidian otomatis 3. bikin dashboard Meridian lebih detail 4. bikin runbook VPS recovery / incident page
• [assistant] ✅ Semua 4 saya jalankan. ## Yang dibuat ### 1) Daily note template ```text Templates/Daily Note Template.md ``` Isi: - Top 3 - Ops Check - Trading / Meridian - Infra - Notes - Next Actions ### 2) Page khusus ops ```text Areas/Operations/Meridian.md Areas/Opera

MCP

• [assistant] ✅ **Normal. Sudah saya test real.** ## Hasil test MCP ```text initialize -> 200 tools/list -> 200 list_notes -> 200 read_note -> 200 ``` ## Data nyata dari vault `list_notes`: ```text welcome.md ``` `read_note welcome.md`: ```md # Welcome to Andy Obsidian Vaul
• [assistant] ✅ **Sudah jalan normal. Full beres.** ## Yang saya lakukan 1. Verifikasi token Cloudflare valid 2. Hapus record lama: ```text CNAME obsidian -> *.cfargotunnel.com ``` 3. Buat record baru: ```text A obsidian -> 43.134.72.148 proxied = false ``` 4. Reload Caddy
• [assistant] TLS belum beres. Root cause baru: ## Sekarang DNS sudah benar Saya sudah ubah sendiri via Cloudflare API: ```text obsidian.andotherstori.my.id A -> 43.134.72.148 proxied = false ``` ## Tapi HTTPS belum punya cert valid Makanya `hermes mcp test` masih timeout.

MERIDIAN

• [assistant] Enhance Meridian dashboard and VPS recovery.
• [assistant] Sync OK. - Added/synced: 145 wallets - Output: `Synced 145 wallets to /home/ubuntu/meridian/dumper-wallets.json` No errors.
• [user] 1. bikin template daily note 2. bikin page khusus Meridian / VPS / Hermes ops 3. import knowledge penting dari session ke Obsidian 4. bikin struktur project yang lebih rapi

VPS

• [assistant] Karena yang kamu kasih **bukan akses DNS Cloudflare**. ## Yang kamu kasih Kamu kasih ini: ```bash cloudflared service install eyJh... ``` Itu = **tunnel token**. Token itu cuma bisa: - nyalain connector `cloudflared` - connect VPS ke tunnel Cloudflare Token it
• [assistant] Itu **nggak perlu**. `cloudflared` di VPS **sudah terinstall**. Bukan itu masalahnya. ## Status sekarang | Item | Status | |---|---| | cloudflared install | ✅ sudah ada | | MCP server | ✅ jalan | | Caddy | ✅ jalan | | DNS/Tunnel routing | ❌ masih salah untuk H
• [assistant] Lakukan ini. Singkat. Exact. ## Yang salah sekarang Kamu bikin: ```text obsidian.andotherstori.my.id CNAME -> 12238757-a771-4e80-b1df-109c1c22546b.cfargotunnel.com ``` Itu masih lewat **Cloudflare Tunnel edge**. Hermes MCP masih macet di jalur itu. ## Yang ben

CRON

• [assistant] Test insight script and maybe create cron jobs.
• [user] [IMPORTANT: You are running as a scheduled cron job. DELIVERY: Your final response will be automatically delivered to the user — do NOT use send_message or try to deliver the output yourself. Just produce your report/output as your final response and the syste
• [assistant] Beres. Gue sudah probe, pilih route yang hidup, patch, restart, dan verify real output. # Hasil probe 9router ## Yang hidup | Model | Status | |---|---| | `combo` | OK | ## Yang gagal | Model | Error | |---|---| | `combo-meridian` | `429 Too Many Requests` | |

INSTAGRAM

• [assistant] Lanjut apa, Andy? Status terakhir: - ✅ enowX live: https://enowx.andotherstori.my.id - ✅ Instagram cron fixed, pipeline OK - ✅ Apify token valid - ✅ VPS disk/RAM safe - ✅ 9router running - ✅ Google Workspace authenticated Pilih next: 1. Setup provider/account
• [assistant] Instagram daily dashboard sudah kembali normal — pipeline 5/5 akun sukses total tanpa error. Tadi sempat *rate limit* karena model `deepseek-v4-flash` habis kuota harian, tapi setelah saya kembalikan ke model `combo` (via 9router), pipeline langsung *green* 10
• [user] Cronjob Response: Instagram daily dashboard automation (job_id: 221bca1b28a3) ------------- EXIT=0. Dashboard run OK. All steps (SocialBlade, Supabase, Apify, Sheet) OK. Git commit/push skipped (no changes detected). Final report: - SocialBlade: ✅ OK (5 accoun

Meridian Project Dashboard

Mission

• Keep bot profitable enough through fee farming, not blind scalp exits
• Reduce bad deploys, avoid instant OOR, protect capital

Current Important Facts

• maxMcap: $1M
• maxTvl: $40K
• minFee: 0.3
• maxVol: 5
• regime guard disabled
• route: combo-meridian OK

Current Weak Spots

• Prompt lacks price-vs-active_bin context
• `entry_token_age_hours` bug stores `***`
• Multi-pool token memory still needs clean base_mint cross-ref confidence

Active Tracking

• entry_token_age_hours
• entry_is_new_ath
• entry_trend_context
• min_pnl_pct
• max_drawdown_pct

Next Fix Candidates

• Add price-vs-active_bin to deploy prompt
• Fix token age numeric storage
• Review OOR exits against productive fee-farm holds

Linked Notes

• [[Areas/Operations/Meridian]]
• [[Inbox/Session Knowledge Import]]

VPS Project Dashboard

Scope

• Stability
• Memory / swap
• Disk cleanup
• PM2 / service health
• DNS / TLS

Current Important Facts

• OOM issue mitigated
• disk cleanup done
• swap expanded
• obsidian MCP live

Links

• [[Areas/Operations/VPS]]

Hermes Project Dashboard

Scope

• MCP integrations
• Cron stability
• Session hygiene
• Agent workflows

Current Important Facts

• Obsidian MCP connected
• Hindsight MCP connected
• Instagram cron restored
• On-this-day moved off insufficient-balance model

Links

• [[Areas/Operations/Hermes]]

Meridian Daily Review — 2026-07-08

Summary

• Mood / regime:
• Net result:
• Main lesson:

Deploy Quality

• Good deploys:
• Bad deploys:
• Instant OOR count:

Exit Review

• Best exit:
• Worst exit:
• Stop-loss notes:

Prompt / Filter Notes

• price-vs-active_bin:
• token_age signal:
• ATH / trend context:

Action Items

• [ ]
• [ ]
• [ ]

Meridian Weekly Review — 2026-W28

Weekly Summary

• Net result:
• Best improvement:
• Biggest mistake:

Strategy Review

• Deploy quality:
• OOR behavior:
• Stop-loss quality:
• Fee farming quality:

Prompt / Filter Review

• price-vs-active_bin:
• token_age signal:
• trend context:
• ATH handling:

Infra / Reliability

• 9router:
• VPS:
• Cron / data feed:

Next Week Actions

• [ ]
• [ ]
• [ ]

VPS Incident Log

Use template: [[Templates/VPS Incident Log Template]]

Latest

• No new incident logged yet