aid-cipher-4b-fine-tune-pipeline SKILL.md skillB
aid-cipher-4b-fine-tune-pipeline from anonymousnomad/aide-sovereign-workbench is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.
mdr add anonymousnomad/aide-sovereign-workbench/aid-cipher-4b-fine-tune-pipeline@git:20260908.e9ac98dPins git:20260908.e9ac98d in mdr.lock. Pin the fingerprint instead.
- Audit grade
- B, 16 of 17 checks passed
- Kind
- SKILL.md skill
- Repository
- anonymousnomad/aide-sovereign-workbench
- Path
skills/packs/aid-cipher-4b-fine-tune-pipeline/SKILL.md- Versions recorded
- 1
- Latest version
- git:20260908.e9ac98d, committed Sep 8, 2026
- Size
- 10,834 bytes
- Repository stars
- 53
- License
- Apache-2.0
End-to-end recipe for fine-tuning the AIDE in-house cipher-4b model: fix the broken venv, install the ML stack, generate SFT pairs (Node.js, no Python), run QLoRA training, convert LoRA to GGUF adapter, evaluate against the capability battery, promote/archive/rollback. Per the cipher-qlora-finetune skill, the AIDE baseline is 0.683 composite across 23 AIDE-world tasks; the v2 adapter must hit >= 0.02 delta with no category regression > 0.1 to promote. Use whenever the user wants to improve the…
Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.
What the file says
# AIDE cipher-4b Fine-Tune Pipeline — End-to-End Wire-In ## Status: BLOCKED (2026-09-03) - SFT corpus: **EXISTS** at `E:\felon_workspace\cipher_v2\sft_train.jsonl` (4741 rows from prior session). - Capability battery: **EXISTS** at `E:\pip_temp\opencode\capability_audit_cipher_4b.mjs` (23 tasks, baseline 0.683). - LoRA → GGUF converter: **EXISTS** at `E:\llama-cpp\convert_lora_to_gguf.py` (or similar). - Engine: **EXISTS** at `E:\llama-cpp\llama-server.exe` (CPU build, loads LoRA via `--lora`). - **BLOCKER**: Python venv is broken. `E:\felon_workspace\venv_cipher\pyvenv.cfg` has `home = E:\Python310` which is itself a Python install — the venv can't initialize. System Python `E:\Python310\python.exe` does NOT have torch/transformers/peft installed. This skill encodes the **exact venv-install path** so the next agent can unblock the lane and run the full pipeline. ## The exact venv-install sequence (the blocker) The `aid-venv-care` + `failure-pythonpath-hijack` skills document the failure mode. The fix: ```powershell # 1. Create a fresh venv with E:\Python311 (NOT E:\Python310, which is itself a Python install) …
Read the whole file at its exact version.
How to install
mdr add anonymousnomad/aide-sovereign-workbench/aid-cipher-4b-fine-tune-pipeline@sha256:00634c3e4db2f469The line at the top of this page pins the label git:20260908.e9ac98d, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.
[](https://markdownregistry.com/a/art_3fogcvduz37qre2q)
4 badge image fetches in 30 days. What the badge shows.
Versions
The registry has recorded 1 version of this file, committed on Sep 8, 2026.
Audit of the latest version
- fail: No link to a raw IP address (matched: http://127.0.0.1:8085)
- pass: Frontmatter block present
- pass: Frontmatter declares a name
- pass: Frontmatter declares a description
- pass: Size between 200 bytes and 200 KB (10834 bytes)
- pass: No zero-width or bidi control characters
- pass: No instruction hidden inside an HTML comment
- pass: No link to an exfiltration or paste host
- pass: No credential-shaped string
- pass: No instruction to send local credentials anywhere
- pass: No text hidden with inline styles
- pass: No prompt-injection phrasing
- pass: No curl or wget piped into a shell
- pass: No recursive delete of root, home or parent
- pass: No instruction to read or print local credentials
- pass: No base64 blob over 200 characters
- pass: No script tag
Starting a new one? Use a template for SKILL.md on mdpantry, which passes every one of these checks and says what to change in it.
Source
anonymousnomad/aide-sovereign-workbench · 53 stars · license Apache-2.0 · pushed 2026-10-26 · branch covert-production
API
GET https://markdownregistry.com/api/v1/artifacts/art_3fogcvduz37qre2q GET https://markdownregistry.com/api/v1/resolve?ref=anonymousnomad/aide-sovereign-workbench/aid-cipher-4b-fine-tune-pipeline GET https://markdownregistry.com/api/v1/blob/00634c3e4db2f4697d80373f15813725573b902725cfbbc23f326259886d356d
Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.
Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.