Analog Extraction Resim
Post-layout parasitic extraction + re-simulation for analog blocks — compares pre-layout vs post-layout specs. Use when the user says "post-layout resim", "extract and resimulate", "parasitic check", or at Step A7 of the analog track.
Get the complete skill folder, including its scripts and reference files.
When to use this skill
Use Analog Extraction Resim when an AI agent needs a reusable SKILL.md workflow for this job: Post-layout parasitic extraction + re-simulation for analog blocks — compares pre-layout vs post-layout specs. Use when the user says "post-layout resim", "extract and resimulate", "parasitic check", or at Step A7 of the analog track.
When not to use it
Skip Analog Extraction Resim when the task is outside the productivity category, or when a more specific skill in this directory already covers the same workflow with clearer triggers.
How to install
Where will you use it?
mkdir -p ~/.claude/skills/analog-extraction-resim # Save SKILL.md to ~/.claude/skills/analog-extraction-resim/SKILL.md
- Claude Code personal install: place the skill and its bundled files at ~/.claude/skills/analog-extraction-resim/SKILL.md. For a shared project, commit the folder at .claude/skills/analog-extraction-resim/ instead.
- Current Claude Code detects SKILL.md changes in watched directories during the session. If you just created a top-level skills directory, run /reload-skills. Check that the skill is listed before testing its trigger.
- For Claude chat or Cowork in the Desktop app, manage uploaded skills in Customize > Skills; copying files into a local Claude Code folder is not a Desktop install. Other agents may use different install locations.
Full install guide for Claude, Cursor, and Codex →
Try a starter prompt
After installing the skill, paste this starter prompt into your agent to verify it triggers on a small task.
Use the Analog Extraction Resim skill for this task: Step A7 of the analog track. Confirm any required inputs or files before making changes.
What supports this recommendation?
81/100 provisional content score. This measures documentation, practical detail and source signals. It is not a measured task success rate or a security certification.
- Documentation
- 92/100
- Practical detail
- 92/100
- Source evidence
- 53/100
- Source signals
- 72/100
Model compatibility evidence
- claudeuntested
No model-specific signal or recorded compatibility test was found.
- chatgptuntested
No model-specific signal or recorded compatibility test was found.
- geminiuntested
No model-specific signal or recorded compatibility test was found.
- copilotuntested
No model-specific signal or recorded compatibility test was found.
- llamauntested
No model-specific signal or recorded compatibility test was found.
- perplexityuntested
No model-specific signal or recorded compatibility test was found.
- mistraluntested
No model-specific signal or recorded compatibility test was found.
- grokuntested
No model-specific signal or recorded compatibility test was found.
Inferred compatibility is based on content signals. Check the app, its available tools and setup requirements before running a skill.
What this skill does
# Analog Extraction Resim
After analog layout in Magic, extracts parasitic RC and re-simulates across PVT corners to check for layout-induced performance degradation. Compares pre-layout vs post-layout results and flags regressions.
## When to use
- Step A7 of the analog track - After `analog-layout` has produced a Magic `.mag` file - When the user asks "did the layout hurt my bandwidth?"
## Inputs
1. `analog/<block>/layout.mag` — Magic layout file 2. `analog/<block>/corner_results.json` — pre-layout SPICE results (baseline) 3. `analog/<block>/spec.json` — specs for pass/fail comparison 4. PDK (gf180 or sky130)
## Workflow
1. **Extract parasitics** — emit/validate the Magic parasitic-RC TCL with `programs/magic_extract_spice_emit.py` (do not hand-write the recipe): ```bash # emit the deterministic .mag -> RC-annotated .subckt extraction TCL python3 programs/magic_extract_spice_emit.py --block <block> \ --out-spice analog/<block>/<block>_extracted.spice --out extract.tcl # or validate an existing extraction TCL (FAILs if it omits `extract all` # or `ext2spice lvs` — the two silent causes of a vacuous 0% degradation) python3 programs/magic_extract_spice_emit.py --validate extract.tcl ``` Then run it via `eda_extraction` (or magic). Output: `analog/<block>/<block>_extracted.spice`. > The fixed `load / extract all / ext2spice lvs / ext2spice` recipe is > enforced by `programs/magic_extract_spice_emit.py` (distinct from the > GDS-read + port-promote LVS recipe in `magic_port_extract_emit.py`).
2. **Re-simulate with extracted netlist**: - Replace ideal subcircuit with extracted netlist in testbench - Run `eda_spice_corner` with same corners as pre-layout - Output: `analog/<block>/post_layout_corner_results.json`
3. **Compare pre vs post** — run the deterministic checker; do not re-grade by hand: ```bash python3 programs/analog_pre_vs_post_layout_check.py <project> --json ``` It computes per-metric per-corner degradation `(post - pre) / pre × 100%` and classifies it against the **canonical degradation bands, which the program owns** (single source of truth — see "Degradation thresholds" below). Do NOT hardcode a different ERROR/WARNING cutoff in your report; quote the program's verdict. - Typical degradation sources: - Bandwidth reduction (parasitic C on high-impedance nodes) - Gain reduction (parasitic R in signal path) - Increased noise (parasitic coupling)
## Output format
### `analog/<block>/pre_vs_post.json` ```json { "block_name": "ldo_1v8", "pre_layout_file": "corner_results.json", "post_layout_file": "post_layout_corner_results.json", "comparison": { "gain_db": {"pre": 62.3, "post": 58.1, "degradation_pct": -6.7, "status": "OK"}, "ugb_mhz": {"pre": 11.2, "post": 7.5, "degradation_pct": -33.0, "status": "ERROR"}, "vout_dc": {"pre": 1.8002, "post": 1.7998, "degradation_pct": -0.02, "status": "OK"} }, "worst_degradation": {"metric": "ugb_mhz", "pct": -33.0}, "overall_status": "NEEDS_RELAYOUT" } ```
## Degradation thresholds
Enforced by `programs/analog_pre_vs_post_layout_check.py` (single source of truth — `≤20%` OK / `>20%` WARNING / `>30%` ERROR→NEEDS_RELAYOUT). Quote the program's verdict; do not restate a conflicting cutoff. If the policy must change, change it in the program (one place) so SKILL.md and runtime never drift.
## Do not
- Do not skip extraction and go straight to hardmacro — parasitic RC is the #1 cause of analog silicon failure - Do not compare only TT corner — worst-case degradation often appears at SS+hot - Do not ignore capacitive loading on compensation nodes (Cc) — parasitics add to Cc
## Handoff
Branch on the `overall_status` field emitted by `programs/analog_pre_vs_post_layout_check.py` (deterministic, not a judgment call):
- `OK` / `WARNING` → `analog-hardmacro-gen` (Step A8) - `NEEDS_RELAYOUT` → back to `analog-layout` (Step A5) - `post_layout_corner_results.json` → `analog_pre_vs_post_layout_check` gate
## Compliance gate (mandatory)
After producing your output, save it to a file and run:
```bash python3 plugins/vibe-ic/_shared/skill_compliance_check.py \ --requirements plugins/vibe-ic/skills/analog-extraction-resim/compliance.yaml \ <your_output_file> ```
Exit 0 = PASS, exit 1 = FAIL with specific missing elements listed.
**Your task is not complete until the audit returns PASS.**
Intended uses
- Step A7 of the analog track
- After analog-layout has produced a Magic .mag file
- When the user asks "did the layout hurt my bandwidth?"
- Extract parasitics — emit/validate the Magic parasitic-RC TCL with
- Re-simulate with extracted netlist:
- Replace ideal subcircuit with extracted netlist in testbench
- Run edaspicecorner with same corners as pre-layout
- Output: analog/<block>/postlayoutcornerresults.json
- Compare pre vs post — run the deterministic checker; do not re-grade by hand:
- Typical degradation sources:
- Bandwidth reduction (parasitic C on high-impedance nodes)
- Gain reduction (parasitic R in signal path)
- Increased noise (parasitic coupling)
Implementation notes
- Review the source instructions and adapt inputs before running the workflow.
Try it, then tell us how it went
A copied prompt does not tell us whether the task worked. Report your result after checking the output. We collect your tool and outcome, never your files or prompt.