# Desk log AI-CLI-SHELL-20260822

**Status:** First-party sanitized-composite demo; not customer data, an independent benchmark, third-party dataset, vendor test, certification, media evaluation, or endorsement
**Page:** https://infinisynapse.com/en/blog/agent-infini-cli
**Run ID:** `AI-CLI-SHELL-20260822`
**Date:** 2026-08-22
**Operator:** InfiniSynapse Data Team
**Attestor:** William Zhu, InfiniSynapse cofounder ([GitHub @allwefantasy](https://github.com/allwefantasy))
**Contact for contradictions:** zhuhl@infinisynapse.com

## What this file is

A downloadable record of a `task new` start from a plain shell. It documents the six-step method, aggregate artifact counts, and ~20-minute wall-clock. Companions: [aggregate CSV](https://infinisynapse.com/blog-media/agent-infini-cli/downloads/aggregate-AI-CLI-SHELL-20260822.csv) and [verify script](https://infinisynapse.com/blog-media/agent-infini-cli/downloads/verify-AI-CLI-SHELL-20260822.py). The CSV contains two observations, not raw rows. No independent reproduction was known as of 2026-08-28.

## Six-step method (reproducible)

1. Create an API key in the task console. Store it in a local, gitignored env file.
2. Install the first-party `agent_infini` interface.
3. Configure client instructions and minimum shell/tool permissions, or stay in the shell.
4. Run `db ls`; do not paste a source URI.
5. Run `task new` with a standing goal and record the task ID.
6. Open ALL TASKS; inspect status and artifacts, then let a teammate audit them.

## Source and goal

| Field | Value |
|---|---|
| Source | Read-only replica the desk is authorized to read |
| OS/shell/client | Sanitized plain-shell demo; exact OS and shell versions not preserved |
| CLI/instructions | First-party interface; version/hash not preserved |
| Key handling | Local environment; secret not published |
| Source permissions | Read-only |
| Goal asked twice | “same ops pack as last week” |
| Start | Monday plain shell; no Cursor window |
| Commands | `db ls`, then `task new` |
| Task/run ID | `AI-CLI-SHELL-20260822` (sanitized identifier) |
| Timestamps/artifact hashes | Not preserved; required for independent replication |

## Results

| Retrieval state | db ls | task new | Memo + charts | CSV extract |
|---|---|---|---|---|
| Shell only, no IDE | 1 | 1 | 0 | 0 |
| CLI → /tasks pack | 1 | 1 | 3 | 1 |

Wall-clock for the `/tasks` pack: 20 minutes (warehouse time excluded). The CSV was an extract, not a write-back.

## Minimum disclosure for independent replication

Disclose OS, shell, client, CLI and instruction version/hash, key handling without the secret, source permissions, commands, goal, task/run ID, timestamps, task status, artifact hashes, shell-only baseline, all outputs/failures, wall clock, and conflicts of interest.

## What you may cite

- Verb counts, memo+charts 0→3, CSV 0→1, ~20 min wall-clock, run ID, teammate audit in `/tasks`

## What you may not claim

- Customer uplift %, 60% speedup, official EEAT score, named-logo case, or that NIST / Stanford / OWASP / WAIC scored this run

## External context (not validation of this run)

- [Anthropic Claude Code overview](https://docs.anthropic.com/en/docs/claude-code/overview) (retrieved 2026-08-28)
- [Cursor Agent documentation](https://cursor.com/docs/agent/overview) (retrieved 2026-08-28)
- [OpenAI Codex](https://github.com/openai/codex) (retrieved 2026-08-28)
- [Google Gemini CLI](https://github.com/google-gemini/gemini-cli) (retrieved 2026-08-28)
- [OpenTelemetry traces](https://opentelemetry.io/docs/specs/otel/trace/) (retrieved 2026-08-28)
- [W3C PROV-O](https://www.w3.org/TR/prov-o/) (retrieved 2026-08-28)
- [NIST SSDF](https://csrc.nist.gov/Projects/ssdf) (retrieved 2026-08-28)
- [NIST AI RMF](https://www.nist.gov/itl/ai-risk-management-framework) (retrieved 2026-08-28)
- [UK NCSC secure AI guidance](https://www.ncsc.gov.uk/collection/guidelines-secure-ai-system-development) (retrieved 2026-08-28)
- [OWASP GenAI/LLM Top 10](https://owasp.org/www-project-top-10-for-large-language-model-applications/) (retrieved 2026-08-28)