Vivado Impl

What it does

When your agent uses it

• The user needs Vivado implementation strategy selection
• Tasks that involve Task breakdown

Example prompts

• “/vivado-impl”

Workflow steps

5 steps, taken from the first numbered list in SKILL.md.

1. 1Synthesis: Use AlternateRoutability directive, reduce MUXF/CARRY usage
2. 2opt_design: Use -muxf_remap, -carry_remap, LUT_DECOMPOSE property
3. 3place_design: Use AltSpreadLogic_high or SSI_SpreadLogic_high
4. 4route_design: Use AlternateCLBRouting
5. 5RTL changes: Reduce fanout, pipeline long paths, reduce resource utilization

What it can do on your machine

Read from SKILL.md and the folder at commit b60a52e. It shows what the files ask for, not the result of running them.

• Tool permissions

  Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

  From allowed-tools in the SKILL.md frontmatter.

• Runs code

  No scripts in the folder and no shell commands in SKILL.md (its code samples are tcl).

  From the folder's file list and the shell code blocks in SKILL.md.

• Network

  No URLs in SKILL.md.

  From URLs in SKILL.md, links to its own repository left out.

• Credentials

  Names no API keys, tokens, secrets or passwords.

  From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Vivado Impl loads about 3.7k tokens when it runs. Until then it costs about 184 tokens; SKILL.md has 1,183 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~184

When it runs · the whole SKILL.md, loaded when a task matches

~3.7k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from Shinei-Nouzen-Arch/FPGA-Agent at commit b60a52e, republished under its GPL-2.0 licence (© Shinei-Nouzen-Arch). 1,183 words, ~3,671 tokens.

name

vivado-impl

description

Use this skill when the user needs Vivado implementation strategy selection or optimization. Covers opt_design, power_opt_design, place_design, phys_opt_design, route_design, implementation directives/options, congestion-driven placement, fanout/placement/routing/SLR/register physical optimization, hold fixing, incremental implementation, ECO flow, and implementation run strategies for performance, congestion, area, and power. This skill chooses implementation-phase tactics and tradeoffs. For TCL command generation/execution use vivado-tcl, for synthesis use vivado-synth, for constraints use vivado-constraints, for report interpretation use vivado-analysis, for end-to-end timing closure use vivado-timing-closure.

opt_design          ← Logic optimization (REQUIRED)
  ↓
power_opt_design    ← Clock gating power opt (OPTIONAL, not Versal)
  ↓
place_design        ← Placement (REQUIRED)
  ↓
power_opt_design    ← Post-place power opt (OPTIONAL)
  ↓
phys_opt_design     ← Post-place physical opt (OPTIONAL, recommended)
  ↓
route_design        ← Routing (REQUIRED)
  ↓
phys_opt_design     ← Post-route physical opt (OPTIONAL)
  ↓
write_bitstream     ← Bitstream (all except Versal)
write_device_image  ← Device image (Versal only)

# Route top 10 critical nets first with minimum delay
set preRoutes [get_nets -of [get_timing_paths -max_paths 10]]
route_design -nets [get_nets $preRoutes] -delay
# Then route rest preserving critical routes
route_design -preserve

# Basic usage (optimize entire design)
power_opt_design

# Control scope
set_power_opt -include_cells [get_cells inst_A]
set_power_opt -exclude_cells [get_cells inst_B]
set_power_opt -cell_types {BRAM}
set_power_opt -clocks {clk1}

# Non-Project Mode
read_checkpoint -incremental <reference_routed.dcp>
# Automatic incremental (recommended)
read_checkpoint -incremental -auto_incremental <reference.dcp>
# Force incremental even if criteria not met
read_checkpoint -incremental -force_incr <reference.dcp>
# Fix specific objects
read_checkpoint -incremental <ref.dcp> -fix_objects [all_rams]

read_checkpoint -incremental -directive <directive> <ref.dcp>

report_incremental_reuse -file incr_reuse.rpt

report_design_analysis -congestion -file congestion.rpt
report_route_status -file route_status.rpt

© Shinei-Nouzen-Arch, GPL-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 8 other files in vivado-impl of Shinei-Nouzen-Arch/FPGA-Agent.

• SKILL.md
• REFERENCE.md
• agents/openai.yaml
• examples/ug906/TIMING-201/after/single_sdp_ram.sv
• examples/ug906/TIMING-201/before/single_sdp_ram.sv
• examples/ug906/TIMING-202/after/wide_mulitplier.sv
• examples/ug906/TIMING-202/before/wide_mulitplier.sv
• examples/ug906/UTIL-203/after/sp_rom.sv
• examples/ug906/UTIL-203/before/sp_rom.v

Open the folder on GitHubat commit b60a52e