Running Strapi locally
| Field | Value |
|---|---|
| Source | Obsidian · Claudian plugin |
| Session | ses_f46ffe44fffetcoU5KmX6JY4Ht |
| Started | 2026-09-19 09:29 |
| Messages | 55 |
| Parts | 270 (text+patch+tool) · 45 reasoning · 53 tool calls |
| Model | opencode:big-pickle |
| Directory | /home/sambasene/obsidian/Vault |
Transcript
Section titled “Transcript”can you run strapi locally
<linked_content path=“local run.md” />
read — {"filePath": "/home/sambasene/obsidian/Vault/local run.md"}
<path>/home/sambasene/obsidian/Vault/local run.md</path><type>file</type><content>1: freecodecamp.org2: the odin project3: appwrite4: misskey5: arterisk6: virtual box - kali - 950 tester@7: idle twilight8: nextcloud9: bagisto10: openclaw11: hermes12: opencode13: nextcloud14: rocket chat15: linux journey16: n8n17:18:19: ****20:21:22: supabase23: monkey type24: moodle25: dartpad26: kaggle27:28: over the wire bandit29: peertube30:31:32: matrix33: element34: strapi35: freepbx36: rust desk37:38:39:40:41: athena vm 950 nine 246 ten42:43:
(End of file - total 43 lines)</content>skill — {"name": "local-app-setup"}
<skill_content name="local-app-setup"># Skill: local-app-setup
# Local Application Setup (with bilingual documentation)
## Overview
When the user says they want to set up, install, configure, or run some app/software/service locally, this skill drives two parallel tracks every time:
1. **Track B (get it running)** — Work with the user to install and run the thing: give exact commands, capture results, troubleshoot.2. **Track D (document it bilingually)** — Create and progressively maintain a set of Obsidian notes: a **French** report, an **English** report, and **side/companion files** (setup guide, config template, results log). All written in the **user's first-person voice**.
Run *
…[truncated]I’ll get Strapi running locally. Let me first check the environment constraints and read the reference for the documentation format.
bash — `node –version 2>/dev/null; npm –version 2>/dev/null; yarn –version 2>/dev/null; pnpm –version 2>/dev/null; echo “—”; docker –version 2>/dev/null; docker ps 2>/dev/null | head -5; echo “—”; wh
…[truncated]`
v24.20.011.19.011.21.0------sambaseneHOME=/home/sambasene---2026-08-30.md2026-09-07.mdAGENTS.mdAppsArtifactsAuto Biography.mdCahier des Charges - PolyPortal.mdC-escape-characters.pngchrome exploration result.mdClaude Code.mdCommission Pedagogique.mdDevOps.mdDocumentationdraftsESPExcalidrawFedora Setup.mdFlutter Installation Process.mdFlutter Installation Process Raw.mdfreeCodeCamp.org.mdFreePBX Setup Commands Results.mdFree pbx setup past commands.mdGoogle main.pngGreatest of All Time - Idle Zelus.mdHermes Setup Results.mdIdle TwilightJava.mdLearnLight (1).pngLight.pngread — {"filePath": "/home/sambasene/.agents/skills/local-app-setup/references/workflow-example.md"}
<path>/home/sambasene/.agents/skills/local-app-setup/references/workflow-example.md</path><type>file</type><content>1: # Worked Example — Asterisk VOIP (the pattern this skill is based on)2:3: This reference mirrors the concrete deliverable set produced for a real local setup4: (Asterisk VOIP PBX on Fedora, 2026-09-02). Use it as the target shape for **any** new5: `local-app-setup` task. See the actual reports in the vault:6: `Documentation/Projects/Asterisk VOIP Setup Report.md` (EN) and7: `Documentation/Projects/Rapport Asterisk VoIP Setup.md` (FR).8:9: ---10:11: ## Deliverable set produced12:13: ```14: Documentation/Projects/15: ├── Asterisk VOIP Setup Report.md <- EN
…[truncated]Good — Node 24 + npm are available. Now let me check whether I can act directly or must delegate to your host terminal (sandbox test), and whether we have network access.
bash — `echo “— can I write to $HOME? —”; touch /home/sambasene/.write_test 2>&1 && echo “YES writable” && rm /home/sambasene/.write_test; echo “— ls $HOME —”; ls -a /home/sambasene | head -40; echo
…[truncated]`
--- can I write to $HOME? ---YES writable--- ls $HOME ---....actual-data.agents.androidAndroidAndroidStudioProjectsApplicationsappwriteassets.bash_history.bash_logout.bash_profile.bashrc.bundle.cache.cert.claude-science.config.copilotcourses.cua-driver.dartServer.dart-tool.denoDesktopdevdevelop.dockerDocuments.dotnetDownloads.dropboxDropbox.dropbox-dist.emulator_console_auth_token.enve.flutter-devtools.gemini.gitconfig--- /tmp writable? ---YES /tmp writable--- npm registry reachable? ---HTTP/2 200date: Sat, 19 Sep 2026 09:30:20 GMTcontent-type: application/json--- docker/podman? ------ psql/mysql? ---/usr/bin/sqlite3--- flatpak sandbox?
…[truncated]Excellent — I can write to $HOME, network works, sqlite3 is present, but no Docker. So I’ll run Strapi natively (Node + SQLite — Strapi’s quickstart default). Let me check the vault conventions first.
bash — `echo “— vault .gitignore —”; cat /home/sambasene/obsidian/Vault/.gitignore 2>/dev/null; echo “— Apps folder —”; ls -la /home/sambasene/obsidian/Vault/Apps 2>/dev/null; echo “— Documentation/
…[truncated]`
--- vault .gitignore ---# Fichiers de configuration spécifiques à la machine.obsidian/workspace.json.obsidian/workspace-mobile.json.idea/.claudian/
# Fichiers systèmeThumbs.db.DS_Storedesktop.ini
# Cache / temp*.tmp~$*
# Config Obsidian locale (machines differentes).obsidian/
# Nested repo: freeCodeCamp clone (tracked by its own git)Learn/freeCodeCamp/--- Apps folder ---total 88drwxr-xr-x. 1 sambasene sambasene 36 Aug 21 19:55 .drwxr-xr-x. 1 sambasene sambasene 2530 Sep 19 09:00 ..-rwxr-xr-x. 1 sambasene sambasene 87022 Jul 30 15:26 favicon.pngdrwxr-xr-x. 1 sambasene sambasene 30 Aug 21 19:11 Library--- Documentation/Projects ---AsteriskDocs Sites - Public & Priva
…[truncated]The pattern is clear: apps live in ~/dev/<name>, reports in Documentation/Projects/ (EN + FR). Let me scaffold Strapi in ~/dev/strapi with SQLite (no Docker needed).
bash — npx --yes create-strapi-app@latest strapi --quickstart --no-run --use-npm 2>&1 | tail -30
Strapi v5.54.0 🚀 Let's create your new project
To deploy your project, create a new project on the Strapi Cloud dashboard.
? Please log in or sign up. (Use arrow keys)❯ Login/Sign up Skip [?25l[7D[7Cnode:internal/readline/interface:573 throw new ERR_USE_AFTER_CLOSE('readline'); ^
Error [ERR_USE_AFTER_CLOSE]: readline was closed at Interface.pause (node:internal/readline/interface:573:13) at PromptUI.close (file:///home/sambasene/.npm/_npx/2d56dd13733e9da7/node_modules/inquirer/lib/ui/baseUI.js:53:13) at PromptUI.onForceClose (file:///home/sambasene/.npm/_npx/2d56dd13733e9da7/node_modules/inquirer/lib/ui/baseUI.js:31:10) at process.emit (node:events:514:28) { code: 'ERR_US
…[truncated]Strapi prompts “Login/Sign up” even with --quickstart, and the prompt crashed without a TTY. Let me check the available flags to skip it.
bash — npx --yes create-strapi-app@latest --help 2>&1
Usage: create-strapi-app [directory] [options]
create a new application
Options: -V, --version output the version number --quickstart Quickstart app creation (deprecated) --no-run Do not start the application after it is created. --ts, --typescript Initialize the project with TypeScript (default) --js, --javascript Initialize the project with Javascript --use-npm Use npm as the project package manager --use-yarn Use yarn as the project package manager --use-pnpm Use pnpm as the project package manager --install Install dependencies --no-install Do not install dependencies --skip-cloud Skip cloud login and project creation --example Use an example app --no-example Do not use
…[truncated]The flag exists: --skip-cloud. Let me clean up the partial folder and re-run with --skip-cloud --non-interactive.
bash — `ls -la /home/sambasene/dev/strapi 2>/dev/null; rm -rf /home/sambasene/dev/strapi; npx –yes create-strapi-app@latest strapi –quickstart –no-run –use-npm –skip-cloud –non-interactive 2>&1 | tail -
…[truncated]`
Strapi v5.54.0 🚀 Let's create your new project
Strapi Creating a new application at /home/sambasene/dev/strapi
deps Installing dependencies with npmnpm error code EALLOWSCRIPTSnpm error --allow-scripts is not allowed in project-scoped installs. Add the entries to the "allowScripts" field in package.json, or to .npmrc, instead.npm error A complete log of this run can be found in: /home/sambasene/.npm/_logs/2026-09-19T09_33_52_066Z-debug-0.log
Error Oh, it seems that you encountered an error while installing dependencies in your project
Don't give up, your project was created correctly
Fix the issues mentioned in the installation errors and try to run the following command:
…[truncated]The project was created but npm blocked install scripts (looks like a hardened npm config on this machine). Let me check the npm config and fix it.
bash — `echo “— ~/.npmrc —”; cat /home/sambasene/.npmrc 2>/dev/null; echo “— global npmrc —”; npm config ls -l 2>/dev/null | grep -iE “allow|ignore|script” ; echo “— npm config list —”; npm config
…[truncated]`
--- ~/.npmrc ---allow-scripts=netlify-cli,esbuild,unix-dgram,sharp--- global npmrc ---allow-directory = "all"allow-file = "all"allow-git = "all"allow-remote = "all"allow-same-version = false; allow-scripts = [""] ; overridden by userallow-scripts-pending = falseallow-scripts-pin = truedangerously-allow-all-scripts = falsedescription = truediff-ignore-all-space = falseforeground-scripts = falseignore-scripts = falsescript-shell = nullstrict-allow-scripts = falsetoken-description = nullallow-scripts = ["netlify-cli,esbuild,unix-dgram,sharp"]--- npm config list ---; "user" config from /home/sambasene/.npmrc
allow-scripts = ["netlify-cli,esbuild,unix-dgram,sharp"]
; node b
…[truncated]Root cause found: your global ~/.npmrc enables npm’s script lockdown (allow-scripts=...), and the Strapi scaffolder passes --allow-scripts which npm 11.19 rejects in project-scoped installs. The clean fix: run npm install directly inside the project and let it honor the allow list. Let me check the generated package.json first, then install.
read — {"filePath": "/home/sambasene/dev/strapi/package.json"}
<path>/home/sambasene/dev/strapi/package.json</path><type>file</type><content>1: {2: "name": "strapi",3: "version": "0.1.0",4: "private": true,5: "description": "A Strapi application",6: "scripts": {7: "build": "strapi build",8: "console": "strapi console",9: "deploy": "strapi deploy",10: "dev": "strapi develop",11: "develop": "strapi develop",12: "start": "strapi start",13: "strapi": "strapi",14: "upgrade": "npx @strapi/upgrade latest",15: "upgrade:dry": "npx @strapi/upgrade latest --dry"16: },17: "dependencies": {18: "@strapi/database": "5.54.0",19: "@strapi/plugin-cloud": "5.54.0",20: "@strapi/plugin-users-permissions": "5.54.0",21: "@strapi/strapi": "5.54.0",22: "
…[truncated]better-sqlite3 (native module, needs prebuild scripts) and sharp/esbuild are the script-having packages. Let me install inside the project — scripts for allow-listed packages run, others get skipped with warnings.
bash — npm install 2>&1 | tail -25
(no output)
<shell_metadata>shell tool terminated command after exceeding timeout 600000 ms. If this command is expected to take longer and is not waiting for interactive input, retry with a larger timeout value in milliseconds.</shell_metadata>The install timed out at 10 min — Strapi has ~1,500 packages. Let me check whether it’s still progressing or stuck.
bash — `echo “— node_modules size —”; du -sh /home/sambasene/dev/strapi/node_modules 2>/dev/null; echo “— count —”; ls /home/sambasene/dev/strapi/node_modules 2>/dev/null | wc -l; echo “— latest npm
…[truncated]`
--- node_modules size ---71M /home/sambasene/dev/strapi/node_modules--- count ---855--- latest npm log ---2026-09-19T09_36_20_008Z-debug-0.log2026-09-19T09_35_07_774Z-debug-0.log2026-09-19T09_35_07_679Z-debug-0.log--- lockfile? ------ running npm procs ---8593 npm install9369 /bin/sh -c echo "--- node_modules size ---"; du -sh /home/sambasene/dev/strapi/node_modules 2>/dev/null; echo "--- count ---"; ls /home/sambasene/dev/strapi/node_modules 2>/dev/null | wc -l; echo "--- latest npm log ---"; ls -t /home/sambasene/.npm/_logs/ | head -3; echo "--- lockfile? ---"; ls -la /home/sambasene/dev/strapi/package-lock.json 2>/dev/null; echo "--- running npm procs ---"; pgrep -af "npm" | he
…[truncated]Good news: npm install (PID 8593) survived the shell timeout and is still running — 855 packages down, lockfile not yet written. Let me wait and poll.
bash — `sleep 240; echo “— still running? —”; pgrep -af “npm install” | head -2; echo “— node_modules size —”; du -sh /home/sambasene/dev/strapi/node_modules 2>/dev/null; echo “— lockfile? —”; ls
…[truncated]`
--- still running? ---9459 /bin/sh -c sleep 240; echo "--- still running? ---"; pgrep -af "npm install" | head -2; echo "--- node_modules size ---"; du -sh /home/sambasene/dev/strapi/node_modules 2>/dev/null; echo "--- lockfile? ---"; ls -la /home/sambasene/dev/strapi/package-lock.json 2>/dev/null || echo "not yet"--- node_modules size ---0 /home/sambasene/dev/strapi/node_modules--- lockfile? ---not yetThe orphaned process got killed and npm rolled back the tree (node_modules now empty). Better approach: launch install detached with nohup into a log, then poll the log.
bash — nohup npm install --foreground-scripts > /tmp/strapi-npm-install.log 2>&1 & echo "started PID $!"
started PID 10228Detached and logging to /tmp/strapi-npm-install.log. Polling progress…
bash — `sleep 180; echo “— proc —”; pgrep -af “npm install” | grep -v pgrep | head -2; echo “— log tail —”; tail -5 /tmp/strapi-npm-install.log; echo “— node_modules —”; du -sh /home/sambasene/dev
…[truncated]`
--- proc ---10228 npm install--- log tail ------ node_modules ---0 /home/sambasene/dev/strapi/node_modulesno lockfile yetStill in the resolution phase (network fetching). Giving it more time…
bash — `sleep 300; echo “— proc —”; pgrep -af “npm install” | grep -v pgrep | head -2; echo “— log tail —”; tail -8 /tmp/strapi-npm-install.log; echo “— node_modules —”; du -sh /home/sambasene/dev
…[truncated]`
--- proc ---10228 npm install--- log tail ------ node_modules ---0 /home/sambasene/dev/strapi/node_modulesno lockfile yet8+ minutes with no visible progress is suspicious. Let me inspect the process state directly.
bash — `ps -o pid,ppid,stat,etime,%cpu,%mem,wchan:20,cmd -p 10228 2>/dev/null; echo “— children —”; pgrep -P 10228 | xargs -r ps -o pid,stat,etime,%cpu,cmd -p 2>/dev/null; echo “— npm cache activity —
…[truncated]`
PID PPID STAT ELAPSED %CPU %MEM WCHAN CMD 10228 1 Sl 09:57 5.0 1.3 ep_poll npm install--- children ------ npm cache activity ---
<shell_metadata>shell tool terminated command after exceeding timeout 120000 ms. If this command is expected to take longer and is not waiting for interactive input, retry with a larger timeout value in milliseconds.</shell_metadata>Install finished (process gone). Checking the result…
bash — `echo “— log tail —”; tail -15 /tmp/strapi-npm-install.log; echo “— lockfile/node_modules —”; ls /home/sambasene/dev/strapi/package-lock.json 2>/dev/null && echo LOCKFILE_OK; du -sh /home/samba
…[truncated]`
--- log tail ------ lockfile/node_modules ---0 /home/sambasene/dev/strapi/node_modules--- better-sqlite3 build present? ------ sharp present? ---ls: cannot access '/home/sambasene/dev/strapi/node_modules/.bin/': No such file or directorySomething’s off — empty log, no lockfile. npm always writes a debug log though. Let me check it.
bash — ls -lt /home/sambasene/.npm/_logs/ | head -5; echo "=== latest debug log tail ==="; f=$(ls -t /home/sambasene/.npm/_logs/*debug* 2>/dev/null | head -1); echo "FILE: $f"; tail -40 "$f"
total 1272-rw-r--r--. 1 sambasene sambasene 46833 Sep 19 10:20 2026-09-19T10_06_26_825Z-debug-0.log-rw-r--r--. 1 sambasene sambasene 1048932 Sep 19 09:59 2026-09-19T09_36_20_008Z-debug-0.log-rw-r--r--. 1 sambasene sambasene 856 Sep 19 09:35 2026-09-19T09_35_07_774Z-debug-0.log-rw-r--r--. 1 sambasene sambasene 861 Sep 19 09:35 2026-09-19T09_35_07_679Z-debug-0.log=== latest debug log tail ===FILE: /home/sambasene/.npm/_logs/2026-09-19T10_06_26_825Z-debug-0.log560 http fetch GET 200 https://registry.npmjs.org/stream-chain 5875ms (cache miss)561 silly packumentCache full:https://registry.npmjs.org/stream-chain set size:undefined disposed:false562 silly fetch manifest react-router@6.30.6
…[truncated]The debug log reveals the real problem: each registry fetch takes 5–140 seconds (e.g. react-is = 144s) — the network to npmjs.org is very slow, so resolution is grinding through ~1,500 metadata fetches. The install also keeps getting killed when my probe commands hit tool timeouts (shared process group). Two fixes: a project .npmrc so native scripts (better-sqlite3) actually run, and setsid to fully detach the install from the shell.
bash — npm config get registry; echo "---"; cat /home/sambasene/dev/strapi/.npmrc 2>/dev/null || echo "no project npmrc"
https://registry.npmjs.org/---no project npmrcwrite — `{“filePath”: “/home/sambasene/dev/strapi/.npmrc”, “content”: “# Strapi project — local dev\n# Native modules that must run install/prebuild scripts.\n# Merged from the global ~/.npmrc allow-scripts li
…[truncated]`
Wrote file successfully.bash — setsid sh -c 'npm install >> /tmp/strapi-npm-install.log 2>&1; echo "EXIT=$?" >> /tmp/strapi-npm-install.log' >/dev/null 2>&1 & sleep 2; pgrep -af "npm install" | grep -v grep | head -2
12547 sh -c npm install >> /tmp/strapi-npm-install.log 2>&1; echo "EXIT=$?" >> /tmp/strapi-npm-install.log12549 npm installInstall is now fully detached (own session — tools can’t kill it). While it grinds through the slow network, I’ll start both Track B docs per the skill. Creating the EN report:
write — `{“content”: “# Strapi Setup Report\n\n- Project: Strapi (headless CMS) — local run\n- Date: 2026-09-19\n- Status: 🟡 In progress — dependencies installing\n- Goal: Run Strapi locally at
…[truncated]`
Wrote file successfully.Now the French report (mirrored):
write — `{“content”: “# Rapport Strapi Setup\n\n- Projet : Strapi (CMS headless) — exécution locale\n- Date : 2026-09-19\n- Statut : 🟡 En cours — installation des dépendances\n- Objectif : Fair
…[truncated]`
Wrote file successfully.Both reports drafted. Now let me check the install progress:
bash — `pgrep -af “npm install” | grep -v grep | head -2 || echo “npm not running”; echo “— log size/tail —”; wc -c /tmp/strapi-npm-install.log; tail -3 /tmp/strapi-npm-install.log; echo “— lockfile —
…[truncated]`
12547 sh -c npm install >> /tmp/strapi-npm-install.log 2>&1; echo "EXIT=$?" >> /tmp/strapi-npm-install.log12549 npm install--- log size/tail ---103814 /tmp/strapi-npm-install.lognpm warn tar TAR_ENTRY_ERROR ENOSPC: no space left on device, open '/home/sambasene/dev/strapi/node_modules/hast-util-whitespace/package.json'npm warn tar TAR_ENTRY_ERROR ENOSPC: no space left on device, open '/home/sambasene/dev/strapi/node_modules/mdast-util-mdxjs-esm/package.json'npm error A complete log of this run can be found in: /home/sambasene/.npm/_logs/2026-09-19T10_06_26_825Z-debug-0.log--- lockfile ---not yet--- modules count ---855New root cause: ENOSPC — disk full during extraction. Let me check disk space:
bash — df -h /home/sambasene /tmp / 2>/dev/null; echo "--- inodes ---"; df -i /home/sambasene 2>/dev/null | head -3
Filesystem Size Used Avail Use% Mounted on/dev/nvme0n1p5 291G 258G 32G 90% /home/sambasenetmpfs 1.6G 214M 1.4G 14% /tmptmpfs 7.7G 104K 7.7G 1% /--- inodes ---Filesystem Inodes IUsed IFree IUse% Mounted on/dev/nvme0n1p5 0 0 0 - /home/sambaseneWait — that ENOSPC error is from the previous run (the log was appended; it references the 10:06 debug log). The current setsid install is still going. But something filled the disk at 855 packages with 32G free — suspicious. Let me investigate the current state:
bash — `echo “— current npm proc —”; ps -o pid,stat,etime,%cpu -p 12549 2>/dev/null || echo “dead”; echo “— node_modules now —”; ls /home/sambasene/dev/strapi/node_modules 2>/dev/null | wc -l; echo “-
…[truncated]`
--- current npm proc --- PID STAT ELAPSED %CPU 12549 Sl 01:08 26.5--- node_modules now ---855--- df now ---/dev/nvme0n1p5 291G 258G 32G 90% /home/sambasene--- newest debug log ---total 5072-rw-r--r--. 1 sambasene sambasene 230108 Sep 19 10:42 2026-09-19T10_41_18_013Z-debug-0.log-rw-r--r--. 1 sambasene sambasene 874 Sep 19 10:37 2026-09-19T10_37_41_929Z-debug-0.log--- what's big in HOME - top 12 ---21G /home/sambasene/VirtualBox VMs15G /home/sambasene/dev15G /home/sambasene/Android12G /home/sambasene/.local12G /home/sambasene/Documents11G /home/sambasene/.gradle7.5G /home/sambasene/.docker6.4G /home/sambasene/.cache5.1G /home/sambasene/.npm2.3G /home/sambasene/develop2.0
…[truncated]Now it makes sense: the current install (PID 12549, 26% CPU, warm cache) is progressing normally with 32G free — the ENOSPC was a transient disk-full event from the earlier orphaned run competing (there are 21G of VirtualBox VMs on this disk). The current run started at 10:41 and is actively reifying. Waiting for extraction to complete…
bash — `sleep 420; echo “— proc —”; ps -o pid,stat,etime,%cpu -p 12549 2>/dev/null || echo “DEAD”; echo “— modules —”; ls /home/sambasene/dev/strapi/node_modules 2>/dev/null | wc -l; echo “— lockfil
…[truncated]`
--- proc --- PID STAT ELAPSED %CPUDEAD--- modules ---855--- lockfile ---not yet--- EXIT marker ---npm error A complete log of this run can be found in: /home/sambasene/.npm/_logs/2026-09-19T10_41_18_013Z-debug-0.logEXIT=1--- df ---/dev/nvme0n1p5 291G 258G 32G 90% /home/sambaseneThe detached run failed too (EXIT=1). Let me read its debug log to find out why:
bash — f=/home/sambasene/.npm/_logs/2026-09-19T10_41_18_013Z-debug-0.log; grep -nE "error|ERR|ENOSPC|EACCES|EBUSY|code " "$f" | tail -30
9762:7446 error start_new_session, process_group)9763:7446 error ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^9764:7446 error File "/usr/lib/python3.14/subprocess.py", line 1875, in _execute_child9765:7446 error self._posix_spawn(args, executable, env, restore_signals, close_fds,9766:7446 error ~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^9767:7446 error p2cread, p2cwrite,9768:7446 error ^^^^^^^^^^^^^^^^^^9769:7446 error c2pread, c2pwrite,9770:7446 error ^^^^^^^^^^^^^^^^^^9771:7446 error errread, errwrite)9772:7446 error ^^^^^^^^^^^^^^^^^^9773:7446 error File "/usr/lib/python3.14/subprocess.py", line 1819, in _posix_spawn9774:7446 error self.pid = os.posix_spawn(execut
…[truncated]The full story: this run completed extraction and ran the install scripts I allowed — but better-sqlite3 had no prebuilt binary for Node 24, fell back to a source compile via node-gyp, and there’s no C compiler (/usr/bin/clang missing). Options: install a compiler, or point node-gyp at an existing gcc. Let me check what’s available:
bash — `echo “— compilers —”; which gcc g++ cc make clang clang++ python3 2>/dev/null; echo “— sudo? —”; sudo -n true 2>&1 && echo “PASSWORDLESS SUDO OK” || echo “sudo needs password / unavailable”; e
…[truncated]`
--- compilers ---/usr/bin/python3--- sudo? ---/bin/sh: line 1: sudo: command not foundsudo needs password / unavailable--- better-sqlite3 prebuild cache ------ node ABI ---137Key insight: no compiler at all, and no sudo (this shell sees a limited /usr — sandbox). So compiling from source is out. Two paths remain: fetch a better-sqlite3 prebuilt binary for Node 24 (ABI 137), or run Strapi on Node 22 LTS (user-space install, no sudo) which definitely has prebuilds. Let me check why prebuild-install failed first:
bash — cd /home/sambasene/dev/strapi/node_modules/better-sqlite3 && timeout 60 npx prebuild-install --verbose 2>&1 | tail -12
at Module.require (node:internal/modules/cjs/loader:1674:12) at require (node:internal/modules/helpers:157:16) at Object.<anonymous> (/home/sambasene/dev/strapi/node_modules/napi-build-utils/index.js:5:11) at Module._compile (node:internal/modules/cjs/loader:1929:14) { code: 'MODULE_NOT_FOUND', requireStack: [ '/home/sambasene/dev/strapi/node_modules/napi-build-utils/index.js', '/home/sambasene/dev/strapi/node_modules/prebuild-install/bin.js' ]}
Node.js v24.20.0node_modules is corrupt from the failed runs (missing deps in prebuild-install’s tree). Cleanest fix: wipe it, run install with Node 22 LTS (prebuilt binaries exist for its ABI — no compiler needed). Downloading Node 22 user-space:
bash — mkdir -p /home/sambasene/.local/opt && curl -s --max-time 30 https://nodejs.org/dist/latest-v22.x/ | grep -oE 'node-v22\.[0-9]+\.[0-9]+-linux-x64\.tar\.xz' | sort -u | head -3
node-v22.23.2-linux-x64.tar.xzbash — `cd /tmp && curl -s –max-time 300 -o node22.tar.xz https://nodejs.org/dist/latest-v22.x/node-v22.23.2-linux-x64.tar.xz && ls -lh node22.tar.xz && mkdir -p /home/sambasene/.local/opt/node22 && tar -xJf
…[truncated]`
-rw-r--r--. 1 sambasene sambasene 30M Sep 19 10:51 node22.tar.xzv22.23.210.9.8Node 22 installed. Now wipe the corrupt node_modules and reinstall fresh with Node 22’s npm (detached, so nothing can kill it):
bash — `rm -rf /home/sambasene/dev/strapi/node_modules && setsid sh -c ’PATH=/home/sambasene/.local/opt/node22/bin:$PATH npm install > /tmp/strapi-npm-install2.log 2>&1; echo “EXIT=$?” >> /tmp/strapi-npm-inst
…[truncated]`
13792 sh -c PATH=/home/sambasene/.local/opt/node22/bin:$PATH npm install > /tmp/strapi-npm-install2.log 2>&1; echo "EXIT=$?" >> /tmp/strapi-npm-install2.log13793 npm install
<shell_metadata>shell tool terminated command after exceeding timeout 30000 ms. If this command is expected to take longer and is not waiting for interactive input, retry with a larger timeout value in milliseconds.</shell_metadata>Install restarted under Node 22 (the tool just waited on the background task’s file descriptors — the processes are detached). Polling:
bash — `sleep 300; echo “— proc —”; pgrep -af “npm install” | grep -v grep | head -1 || echo DEAD; echo “— log —”; wc -c /tmp/strapi-npm-install2.log; tail -3 /tmp/strapi-npm-install2.log; echo “— m
…[truncated]`
--- proc ------ log ---1308 /tmp/strapi-npm-install2.log
Run `npm audit` for details.EXIT=0--- modules/lockfile ---856LOCK_OKInstall succeeded (EXIT=0) — 856 modules, lockfile written. Now verify the native binaries are real and start the server:
bash — `echo “— better-sqlite3 native binding —”; ls -la /home/sambasene/dev/strapi/node_modules/better-sqlite3/build/Release/*.node 2>/dev/null || echo MISSING; echo “— strapi bin —”; ls /home/sambas
…[truncated]`
--- better-sqlite3 native binding ----rwxr-xr-x. 1 sambasene sambasene 2095488 Mar 13 2026 /home/sambasene/dev/strapi/node_modules/better-sqlite3/build/Release/better_sqlite3.node--- strapi bin ---strapi--- quicksand? quick test require ---better-sqlite3 OK, sqlite 3.51.3Native SQLite works. Starting the Strapi dev server (first boot creates the DB and builds the admin panel):
bash — `rm -f /tmp/strapi-dev.log; setsid sh -c ‘PATH=/home/sambasene/.local/opt/node22/bin:$PATH npm run develop > /tmp/strapi-dev.log 2>&1; echo “DEVEXIT=$?” >> /tmp/strapi-dev.log’ >/dev/null 2>&1 & sleep
…[truncated]`
launchedbash — `sleep 45; echo “— dev log —”; tail -20 /tmp/strapi-dev.log; echo “— port 1337? —”; (timeout 5 curl -s -o /dev/null -w “HTTP %{http_code}\n” http://localhost:1337/ 2>&1 || echo “no response yet
…[truncated]`
--- dev log ---npm error code ENOENTnpm error syscall opennpm error path /home/sambasene/obsidian/Vault/package.jsonnpm error errno -2npm error enoent Could not read package.json: Error: ENOENT: no such file or directory, open '/home/sambasene/obsidian/Vault/package.json'npm error enoent This is related to npm not being able to find a file.npm error enoentnpm error A complete log of this run can be found in: /home/sambasene/.npm/_logs/2026-09-19T10_59_23_132Z-debug-0.logDEVEXIT=254--- port 1337? ---no response yetClassic mistake on my part — the launch command didn’t use the strapi working directory, so npm looked at the vault root. Relaunching with an explicit cd:
bash — `rm -f /tmp/strapi-dev.log; setsid sh -c ’cd /home/sambasene/dev/strapi && PATH=/home/sambasene/.local/opt/node22/bin:$PATH npm run develop > /tmp/strapi-dev.log 2>&1; echo “DEVEXIT=$?” >> /tmp/strapi-
…[truncated]`
relaunchedbash — sleep 60; echo "--- dev log tail ---"; tail -25 /tmp/strapi-dev.log; echo "--- port ---"; timeout 6 curl -s -o /dev/null -w "HTTP %{http_code}\n" http://localhost:1337/ 2>&1 || echo "no response"
--- dev log tail ---✔ Compiling TS (2831ms)
Project information
╭────────────────────┬──────────────────────────────────────────────────╮│ Time │ Sat Sep 19 2026 11:03:07 GMT+0000 (Greenwich Me… ││ Launched in │ 7642 ms ││ Environment │ development ││ Process PID │ 15564 ││ Version │ 5.54.0 (node v22.23.2) ││ Plan │ Community ││ Database │ sqlite ││ Database name │ .tmp/data.db │╰────────────────────┴──────────────────────────────────────────────────╯
Actions available
One more thing...Create your first administrator 💻 by going to the administration panel at:
┌─────────────────────────────┐│ http://localhost:1337/admin │└─────────────────────────────┘
[2026-09-19 11:03:0
…[truncated]🎉 Strapi is live on http://localhost:1337 — started in 7.6s, SQLite DB created. Port responds with HTTP 302 (redirect to admin). Let me verify the redirect target and register the first admin user so it’s fully ready:
bash — `echo “— redirect target —”; timeout 6 curl -s -o /dev/null -w “%{redirect_url}\n” http://localhost:1337/; echo “— register admin —”; timeout 20 curl -s -X POST http://localhost:1337/admin/regi
…[truncated]`
--- redirect target ------ register admin ---Empty responses — let me check if the server is still alive:
bash — `echo “— proc —”; pgrep -af “strapi develop” | grep -v grep | head -2 || echo “strapi DEAD”; echo “— log tail —”; tail -5 /tmp/strapi-dev.log; echo “— direct curl with timing —”; timeout 10
…[truncated]`
--- proc ---15528 node /home/sambasene/dev/strapi/node_modules/.bin/strapi develop15564 /home/sambasene/.local/opt/node22/bin/node /home/sambasene/dev/strapi/node_modules/.bin/strapi develop--- log tail ---└─────────────────────────────┘
[2026-09-19 11:03:07.665] info: Strapi started successfully[2026-09-19 11:03:58.985] http: GET / (855 ms) 302[2026-09-19 11:05:49.428] http: POST /admin/register-admin (49309 ms) 200--- direct curl with timing ---< Date: Sat, 19 Sep 2026 11:07:58 GMT< Connection: keep-alive< Keep-Alive: timeout=5<{ [22 bytes data]* Connection #0 to host localhost:1337 left intactServer is fine — the log shows POST /admin/register-admin … 200: admin user was created (it took 49s on first request while the admin panel was compiling, which is why my curl looked empty). Verifying login works:
bash — `timeout 30 curl -s -X POST http://localhost:1337/admin/login -H “Content-Type: application/json” -d ‘{“email”:“admin@strapi.local”,“password”:“Strapi2026!Local”}’ -o /tmp/login.json -w “HTTP %{http_co
…[truncated]`
HTTP 200--- body ---{"data":{"token":"eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJ1c2VySWQiOiIxIiwic2Vzc2lvbklkIjoiMTY2YmFlMmM0MzI4ZTk5ODQxZGFiMmYwMDJhMDBiODMiLCJ0eXBlIjoiYWNjZXNzIiwiaWF0IjoxNzg5ODE2MTEzLCJleHAiOjE3ODk4MTc5MTN9.ZgrlMTupEnkkDC6-n3CuL2SHi4faEUz3Tg_T6JfBr5s","accessToken":"eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9Login verified — JWT token returned. Strapi is fully operational. Let me update both reports to final status:
edit — {"filePath": "/home/sambasene/obsidian/Vault/Documentation/Projects/Strapi Setup Report.md", "oldString": "- **Status**: 🟡 In progress — dependencies installing\n- **Goal**: Run Strapi locally at htt
…[truncated]`
Edit applied successfully.edit — `{“filePath”: “/home/sambasene/obsidian/Vault/Documentation/Projects/Strapi Setup Report.md”, “newString”: “### Phase 3 — Fix the native module, then start\n\n❌ Mistake 4 / root cause: with the all
…[truncated]`
Edit applied successfully.edit — {"filePath": "/home/sambasene/obsidian/Vault/Documentation/Projects/Strapi Setup Report.md", "newString": "5. **Native modules silently install without their binary when scripts are\n blocked** (bett
…[truncated]`
Edit applied successfully.Now the same updates in French:
edit — `{“filePath”: “/home/sambasene/obsidian/Vault/Documentation/Projects/Rapport Strapi Setup.md”, “newString”: “- Statut : ✅ Terminé — Strapi tourne en local (admin créé, connexion vérifiée)\n- **Obje
…[truncated]`
Edit applied successfully.edit — `{“filePath”: “/home/sambasene/obsidian/Vault/Documentation/Projects/Rapport Strapi Setup.md”, “newString”: “### Phase 3 — Réparer le module natif, puis démarrer\n\n❌ Erreur 4 / cause racine : avec
…[truncated]`
Edit applied successfully.edit — `{“filePath”: “/home/sambasene/obsidian/Vault/Documentation/Projects/Rapport Strapi Setup.md”, “newString”: “5. **Les modules natifs s’installent silencieusement sans leur binaire quand les\n scripts s
…[truncated]`
Edit applied successfully.Now the setup guide side file (reproducible steps, matching the vault’s existing setup-guide pattern):
write — `{“content”: “# Strapi - Setup Guide\n\nHeadless CMS (v5.54.0) running locally on Node 22 + SQLite — no Docker, no DB server.\n\nFinal result: http://localhost:1337/admin\n\n---\n\n## Prerequis
…[truncated]`
Wrote file successfully.Final health check before reporting back:
bash — `echo “— process —”; pgrep -af “strapi develop” | grep -v grep | wc -l; echo “— http check —”; timeout 8 curl -s -o /dev/null -w “GET / → HTTP %{http_code} (%{redirect_url})\n” http://localhost
…[truncated]`
--- process ---2--- http check ---GET / → HTTP 302 (http://localhost:1337/admin)--- db file ----rw-r--r--. 1 sambasene sambasene 933888 Sep 19 11:08 /home/sambasene/dev/strapi/.tmp/data.db--- vault docs ----rw-r--r--. 1 sambasene sambasene 10717 Sep 19 11:11 Rapport Strapi Setup.md-rw-r--r--. 1 sambasene sambasene 3694 Sep 19 11:11 Strapi - Setup Guide.md-rw-r--r--. 1 sambasene sambasene 9736 Sep 19 11:09 Strapi Setup Report.md✅ Strapi is running locally. Open http://localhost:1337/admin and log in with:
admin@strapi.local |
|
| Password | Strapi2026!Local |
What’s running
Section titled “What’s running”- Strapi v5.54.0 on Node v22.23.2 + SQLite (no Docker needed) — PID 15564, DB at
~/dev/strapi/.tmp/data.db - Verified end-to-end:
GET /→ 302 to/admin, admin user registered, login returns HTTP 200 + JWT
The bumpy road (all documented in the vault)
Section titled “The bumpy road (all documented in the vault)”This was a real fight — 4 distinct failures, each with a root cause:
- Strapi v5 prompts for a cloud login during scaffold → fixed with
--skip-cloud --non-interactive - Your global
~/.npmrcscript lockdown made the scaffolder’snpm installdie withEALLOWSCRIPTS(npm 11 forbids--allow-scriptsin projects) → rannpm installmyself with a project-level.npmrc - The npm registry here is brutally slow (5–140 s per fetch) + an ENOSPC transient → detached the install with
setsidso nothing could kill it - No C compiler +
better-sqlite3has no prebuilt for Node 24 → installed Node 22 LTS user-space (~/.local/opt/node22, no sudo) and its prebuilt binary worked immediately ✅
Vault docs created
Section titled “Vault docs created”- [[Documentation/Projects/Strapi Setup Report.md]] (EN, first-person process log)
- [[Documentation/Projects/Rapport Strapi Setup.md]] (FR, mirrored)
- [[Documentation/Projects/Strapi - Setup Guide.md]] (reproducible steps + troubleshooting table)
- Restart any time:
cd ~/dev/strapi && PATH=~/.local/opt/node22/bin:$PATH npm run develop(kill PID frompgrep -af "strapi develop"to stop) - The system Node 24 can’t run it — always use the Node 22 path for this project
Strapi(#34 on yourlocal run.mdlist) is now done — want me to tick it in the list?
what are the default login ?
The login I created during setup is:
| URL | http://localhost:1337/admin |
admin@strapi.local |
|
| Password | Strapi2026!Local |
Let me quickly confirm the server is still up and the credentials still work:
bash — `pgrep -af “strapi develop” | grep -v grep | wc -l; timeout 20 curl -s -X POST http://localhost:1337/admin/login -H “Content-Type: application/json” -d ’{“email”:“admin@strapi.local”,“password”:“Strapi
…[truncated]`
2login → HTTP 200✅ Confirmed working right now — login returns HTTP 200.
- Email:
admin@strapi.local - Password:
Strapi2026!Local
⚠️ One note: that password is the one I chose during setup — it’s fine for local dev, but change it (top-right avatar → Profile) if you ever expose the server beyond localhost. Both reports also hold this in the credentials table ([[Documentation/Projects/Strapi Setup Report.md]]).
45 internal reasoning block(s) omitted — raw thinking is not exported (see archive index).