Skip to content
  1. Skills at a glance
  2. Design and review
  3. /pstack:architect

/pstack:architect

Auto

In one lineBefore code that spans modules, it settles how it's called and what the types are, then compares options.

When to use
The change reaches past one function and touches several modules.
Install with
arena, how
Say it like this
/pstack:architect Design the import flow first, don't jump into code. What I care about most is how other code calls it.

Just asking vs. using it

The same request, typed plainly or with /pstack:architect. Here's what changes.

Just askingWith /pstack:architect
Starts writing the implementation straight away.Sketches types, signatures and module boundaries first, with not implemented bodies.
Runs with the first design it thinks of.Uses arena for sketches from several models, at least two structurally distinct.
Bolts workarounds onto a design that doesn't fit.Reports every deviation from the sketch and scraps it when the same friction repeats.

What the document says

7 sections, about 800 words, 4 min to read. Click a section to read it.

OpeningIt sketches types, signatures, classes and module boundaries with empty bodies, merges several models' designs, then fills in code against the sketch.

  1. 1StartOpens a todolist with one entry each for Ground, Sketch, Agree, Implement and Scrap.
  2. 2Phase A: Ground the problemRuns how on every system the code touches, and why if ownership or layering changes. Skips only for greenfield work.
  3. 3Phase B: SketchRuns arena for sketches, needs two distinct shapes, screens them against the red flags, and prefers a smaller, deeper interface.
  4. 4Phase C: Agree (opt-in)Goes straight to code by default. Pauses for sign-off only on "with checkpoint". Pushback sends it back to A and B.
  5. 5Phase D: Implement against the sketchReplaces not implemented with code. The sketch is the contract, so an unplanned parameter means asking what was wrong.
  6. 6Phase E: Scrap when the architecture is wrongScraps the sketch when workarounds repeat or types need any, subtracts before adding, then reruns arena.
  7. 7OutputsWrites the caller's usage first, derives the type sketch from it, and ships a templated rationale with the synthesis decision.

See the wording worth borrowingRead the full text

Wording worth borrowing

A few sentences from the original: what each one makes the AI do, why it works, and how you can borrow it.

  1. Require a second design

    Require at least two structurally distinct candidates before synthesis, even when the first looks sufficient.

    Makes the AI
    It compares at least two designs with different shapes instead of running with its first idea.
    Why it works
    "Even when the first looks sufficient" blocks the AI's favorite excuse in advance.
    Borrow it
    When you ask for a headline: "Give me two with different angles, even if you like the first."
  2. Design for the laziest successor

    sees only the files it opened, copies the nearest example, and takes the shortest path that compiles

    Makes the AI
    It picks designs that stay right even when someone just copies the nearest example.
    Why it works
    Spelling out the successor's shortcuts as three concrete moves gives the AI something to test each design against.
    Borrow it
    For a layout, write: "Assume the reader sees only the first row and the bold words, and lay it out for that."
  3. Call deviations signals

    Deviations from it are signal to surface, not friction to absorb silently.

    Makes the AI
    When the code stops matching the design, it tells you instead of quietly working around it.
    Why it works
    Recasting "quietly fixing it" as "hiding a signal" tells the AI that small mismatches are worth reporting.
    Borrow it
    When the AI writes from your outline, add: "Where the outline seems wrong, flag it instead of changing it."
  4. List the signs to quit

    The signal is a *pattern*, not single instances.

    Makes the AI
    It throws out the design only when the same trouble keeps coming back, not over one or two odd cases.
    Why it works
    "When to give up" is hard to judge; a list of signs plus "one case isn't enough" keeps it from quitting too early or too late.
    Borrow it
    Set a bar for changing a process: "Change it only if the same complaint comes up three times in a week; log one-offs."
  5. State the default, then exceptions

    Default: proceed directly to implementation with the synthesized design.

    Makes the AI
    By default it builds right after designing, and stops for your sign-off only when you say "with checkpoint".
    Why it works
    A fixed default plus the exact words that trigger the exception means the AI never has to guess whether to pause.
    Borrow it
    When you hand off writing: "By default, send it when done; if I say 'show me first', stop and wait."

Full text

The SKILL.md the AI reads, as written.

descriptionSketch types, signatures, and module structure before code, then stay in the loop while implementation fills in. Use for /pstack:architect, 'architect this', 'design this', or non-trivial work where jumping to code would lock in the wrong shape.

Architect

Sketch types, function signatures, class shapes, and module boundaries with not implemented bodies and pseudocode. Synthesize across multiple model perspectives, then fill in code against the chosen sketch.

Start

Open a todolist with one entry per phase before starting.

  1. Ground
  2. Sketch
  3. Agree
  4. Implement
  5. Scrap

Phase A: Ground the problem

Build a real mental model of every system the new code touches. Run the how skill over the relevant subsystems. Naming a file isn't grounding. Produce the architectural explanation how prescribes. If the design redefines ownership or layering, also run the why skill on the existing shape so the rationale becomes a constraint, not a guess.

Skip Phase A only when the work is genuinely greenfield with no surrounding system to integrate.

Phase B: Sketch

Run the arena skill with the design-sketch task and the Phase A grounding artifacts. Pass references/runner-prompt.md as each runner's prompt.

Take the runners from the architect runners line in the pstack settings file (~/.cursor/rules/pstack-models.mdc in Cursor, ~/.claude/rules/pstack-models.md in Claude Code, ~/.agents/pstack-models.md in other harnesses), in place of the arena runners line. If the file or that line is missing, use claude-opus-5-5-xhigh and grok-4.7-xhigh-fast. Alias and rejected entries follow the runner rules in the arena skill's Phase A.

Design it twice. Require at least two structurally distinct candidates before synthesis, even when the first looks sufficient. This is the exhaust-the-design-space principle skill made concrete. Whole-shape alternatives, not point fixes inside one shape.

Screen every candidate against references/design-red-flags.md before synthesis. Assume the next contributor is an agent that sees only the files it opened, copies the nearest example, and takes the shortest path that compiles. Prefer the design where a change that looks right from one file is right for the whole repo.

Compare viable candidates on interface depth. Prefer the design that hides more complexity behind a smaller, simpler public surface. A rich interface can keep call chains short by concentrating capability instead of scattering it across layers.

Arena returns one synthesized design package. Its synthesis decision fills the rationale's "Synthesis decision" section.

Phase C: Agree (opt-in)

Default: proceed directly to implementation with the synthesized design. No human checkpoint.

Opt in to a checkpoint when the invoker explicitly asks: "/pstack:architect with checkpoint," "stop and show me before implementing," or similar. Then surface the synthesized design and pause for sign-off.

The synthesis can ship as its own commit either way, as the "scaffold first" mode of the foundational-thinking principle skill. Planned and scoped breakage during fill-in is fine, per the outcome-oriented-execution principle skill. For adversarial pressure on the design before implementing, run the interrogate skill on the synthesized sketch.

If the human pushes back on the shape (in a checkpoint or after the fact), treat that as Phase A evidence. Re-ground and re-run Phase B before writing more code.

Phase D: Implement against the sketch

Replace not implemented bodies with code, pseudocode with logic. The synthesized sketch is the contract. Deviations from it are signal to surface, not friction to absorb silently. If a function needs a parameter the sketch didn't anticipate, ask whether the sketch was wrong, the requirement was missed, or the implementation is overreaching.

Phase E: Scrap when the architecture is wrong

If implementation keeps producing friction the sketch can't absorb, throw the sketch out. Don't bolt fixes onto a wrong design, per the redesign-from-first-principles and fix-root-causes principle skills.

The signal is a pattern, not single instances. Tells:

  • The same shape of workaround appearing repeatedly across unrelated code.
  • Multiple unrelated edge cases that all need special-case branches.
  • Types that need escape hatches (any, casts, optional fields always set in practice) to compile.
  • The "we need a lock" reflex when the sketch said the state wasn't shared.
  • Callers having to know the abstraction's internal rules to use it.
  • Two or more independent Phase D deviations of the same shape across the implementation.

Use judgment. A few edge cases don't condemn an architecture. Some problems are legitimately complex. Complexity in the data is not complexity in the design.

When you scrap:

  1. Re-run the how skill over what's been built.
  2. Redesign as if the new constraints had been day-one assumptions, per redesign-from-first-principles.
  3. Subtract before adding, per the subtract-before-you-add principle skill. The new sketch should be smaller than the old one before it grows.
  4. Return to Phase B and re-run arena.

Outputs

The caller's usage is written first and the type sketch derived from it. One file with new types and signatures for small changes. Module map plus type definitions for larger work. The rationale ships alongside, shaped per references/rationale-template.md, including the usage sketch and the synthesis decision.

Original by Lauren Tan (poteto), MIT license. Checked against the pstack mirror at 97de40d. View on GitHub

An unofficial beginner's guide to pstack, not its official docs. pstack is made by poteto; the official repo is the source of truth for commands and behavior. The original guide source is on GitHub.

Hosted by airali. Adapted from Xiaoxiaodong's MIT-licensed guide.