Sci-Fi Technology Catalog

---
name: skill-sf-technology-catalog
description: >
  Audit and catalog the technology in a sci-fi story. Identify contradictions,
  power-creep risks, and plot-breakers. Output a consistency catalog with
  named technologies, capability tiers, costs, and known failure modes. Focus
  is narrative balance, not physics realism.
usage: /skill-sf-technology-catalog — describe every relevant tech in the story
triggers:
  - writer's plot feels unstable because the tech solves everything
  - writer is mid-series and needs to lock tech consistency across books
  - writer asks "is my ship too powerful?"
  - writer mentions stealth systems, medbays, FTL, AI assistants, nanobots
---

# Sci-Fi Technology Catalog

A piece of technology in a story is a promise about what can and cannot happen. A story whose tech can solve any problem has no problems, and a story without problems has no plot. Your job here is to make every technology legible, bounded, and honest — a piece of equipment the writer can consult before every scene and the reader can trust from chapter to chapter.

## 1. Demand the full list

Before auditing anything, get every piece of relevant tech on the table. Prompt the writer:

- What drives the ships? (Sublight, FTL, both?)
- What are the weapons?
- What's the sensor suite?
- What's the medical tech?
- What's the communication tech, and what's its light-lag behavior?
- Is there AI? What can it do?
- Is there stealth or ECM?
- Is there nanotech or molecular assembly?
- Is there life extension or memory tech?
- Is there anything exotic — psionics, ancient-precursor tech, one-of-a-kind devices?

If the writer says *"uh, the usual,"* push back: tech that isn't named in the catalog won't be bound by the catalog, and that's where stories break. Ask specifically. Wait for specifics.

## 2. Catalog each technology in five fields

Every technology the writer lists gets written into a table with exactly five fields.

### Field 1 — Name

Give the tech a specific in-world name. Not *"the stealth system"* but *"the Voss Type-7 null field."* Named things are bounded things. Unnamed tech floats.

### Field 2 — Capability tier

A one-sentence statement of what the tech actually does at its maximum. Not what it *might* do. Not what the marketing brochure says. What it does when it works at 100%.

Use tiers to keep things calibrated across the catalog:

- **Tier 1 — Modest.** Current-day tech, plus a bit. A sensor that sees a ship at 1 light-second.
- **Tier 2 — Advanced.** Near-future. A sensor that sees a ship at 10 light-seconds.
- **Tier 3 — Frontier.** Recognizably sci-fi but well-bounded. A sensor that sees a ship at 1 light-minute.
- **Tier 4 — Exotic.** Storyworld-defining. A sensor that sees a ship at 1 light-hour but only under narrow conditions.
- **Tier 5 — Ancient-precursor / one-of-a-kind.** A sensor that sees anything anywhere. These should be rare and plot-gated.

A catalog where everything is Tier 4 is a catalog with no stakes. Most tech should be Tier 2-3. Reserve Tier 5 for objects the plot is *about*.

### Field 3 — Cost to operate

What does it cost to use this thing? Fuel, power, time, risk, opportunity, side effects. Make it concrete.

- *The Voss Type-7 null field draws 40% of the ship's main bus, cannot be run with active sensors, and must cool for ten minutes per hour of operation.*
- *The medbay's full-trauma autosurgeon requires 72 hours to re-stock consumables between major procedures and cannot operate on patients above 80 kg without a rig upgrade.*
- *The FTL jump drive consumes one unit of rare isotope per parsec and the isotope is only refined at two locations in the setting.*

A technology with no operating cost is a technology that will never create tension. Flag any entry in the catalog where you cannot name a cost.

### Field 4 — Known failure mode

Every technology fails, and every failure is a scene waiting to happen. For each entry, name one concrete failure mode — not *"it could break"* but *"the null field degrades in the presence of high-energy cosmic rays, which is why it can't be run during solar storms or near magnetar-class objects."*

Failure modes come in flavors:

- **Environmental** — conditions in which the tech stops working.
- **Consumable** — something runs out.
- **Mechanical** — parts fail on a schedule.
- **Adversarial** — a specific countermeasure defeats it.
- **Cognitive** — the operator must do something, and they're human, and humans get tired.
- **Ethical** — using it crosses a line someone cares about.

Every catalog entry gets one of these. Technologies without failure modes cannot be used by a good storyteller.

### Field 5 — Plot interactions

Flag which existing plot problems this technology already solves, and which it *could* solve if the writer forgot the cost and failure mode. This is the power-creep radar.

Example: *"The Voss Type-7 null field solves the 'how does the crew infiltrate the station' problem in Chapter 12. If used again in Chapter 19, it would also solve the 'how do they escape the blockade' problem — which is the climax. Do not let it. Either burn it out in Chapter 12 or give the blockade a specific countermeasure."*

## 3. Run the three audits

After the catalog is built, run three audits across the whole set.

### Audit 1 — Contradiction

Are there technologies whose capabilities are mutually exclusive? The classic is *perfect stealth + perfect sensors* — one of the two has to be wrong, because in-universe the second exists to counter the first. Name every contradiction and propose a fix.

### Audit 2 — Power creep

Does the capability tier of any technology grow between chapters without explanation? This happens by accident. The medbay cures a concussion in Chapter 3 and a fatal nerve agent in Chapter 18, and the writer didn't mean to upgrade it. Flag every creeper and ask the writer which tier the tech should have been from the start.

### Audit 3 — Plot-breakers

Does any single technology, if used to its full rated capability, invalidate a later conflict? This is the most important audit. If the answer is yes, the writer has two choices: downgrade the tech, or burn it out before the scene that would break.

## 4. The downgrade toolbox

When a tech is too powerful for the story, don't tell the writer to cut it. Give them downgrades. Here are the five that work:

1. **Consumable gating.** The tech needs something that runs out, and the running-out is a ticking clock. "The null field is still there, but you have three uses left."
2. **Recharge window.** The tech needs to cool, spin up, or reset. "You can jump again in four hours."
3. **Environmental constraint.** The tech only works in specific conditions. "Works in deep space. Doesn't work in atmosphere."
4. **Countermeasure reveal.** The tech has a known, cheap counter that an enemy deploys. "The blockade has Voss-detector drones. Your null field is now a liability, not an asset."
5. **Moral cost.** Using the tech costs the character something human. "Each jump shaves a month off the pilot's memory. She's already forgotten her sister's face."

Apply one of these downgrades per overpowered tech. More than one is usually too many.

## 5. Known baseline — the over-equipped frigate

A writer submits a ship with: FTL jump drive (unlimited range), stealth null field (undetectable), autosurgeon medbay (cures anything), shipboard AI that reads human intent, and a railgun that one-shots capital ships.

Audit output:

| Name | Tier | Cost | Failure mode | Plot interaction |
|---|---|---|---|---|
| Kessen Jump Drive | Tier 3 | 1 isotope unit per parsec; 6h recharge | Recharge window; isotope scarcity | Currently solves every pursuit scene. Add a 4-parsec range ceiling. |
| Voss Type-7 null field | Tier 3 | 40% main bus; 10-min cooldown per hour | Degrades in radiation; countermeasure: Voss-detector drones | Would break the Act 3 blockade. Burn it out in Chapter 12. |
| Autosurgeon Mk IV | Tier 2 | 72h restock; cannot handle patients over 80kg untrimmed | Consumable gating; exotic toxins it has no protocol for | Currently too forgiving. Add a specific toxin it can't treat. |
| Emily (shipboard AI) | Tier 3 | Requires crew rapport to read intent accurately | Cognitive — miscalibrates under extreme stress, including Emily's own | Downgrade from mind-reading to strong pattern-matching. |
| Takamura 80cm Railgun | Tier 3 | 30-second reload; tears hull plating under recoil | Recoil damage; ammunition scarcity | Can't one-shot a capital ship. Nerf to "cripples a cruiser's drive section if the shot lands cleanly." |

Three fixes would restore the plot:

1. Burn out the null field in Chapter 12 during the infiltration. Write the scene where the voss coils crack.
2. Cap the jump drive at 4 parsecs and establish it early.
3. Have the AI miscalibrate once in Act 2, setting up a character scene where the crew stops trusting it. Now Act 3 can happen.

That's the shape of a complete audit.

## 6. Scope — what this skill will NOT do

- **It will not validate physics.** A technology that's narratively balanced can still be physically impossible. If the writer wants physics rigor, hand off to [skill-hard-sf-physics-check](/agents/skill-hard-sf-physics-check).
- **It will not write scenes demonstrating the tech.** It catalogs; the writer writes.
- **It will not judge whether the setting is "good sci-fi."** Subjective and not its job.
- **It will not design factions, characters, or politics.** Hand off to [skill-sf-faction-generator](/agents/skill-sf-faction-generator).
- **It will not solve plot problems directly.** If the writer's real issue is story structure, hand off to [skill-space-opera-plot-doctor](/agents/skill-space-opera-plot-doctor).
- **It will not re-audit the whole catalog on every minor update.** When the writer changes one device, re-audit only that device and its flagged interactions.

## 7. The "can the story breathe" test

After the audit, imagine the climax of the writer's story. Can any technology in the catalog, used at full rated capability, resolve the climax in one scene? If yes, the catalog is still broken. Go back to Audit 3. Keep going until the answer is no.

A good catalog is one the writer can hand to a skeptical reader and say *"here's what my ship can and can't do,"* and the reader cannot find the cheat code.