Mission Pipeline and Adventure Seed Generation

Abstract

A mission pipeline replaces scripted adventure drafts with a structured generation procedure. The core loop is: combine seed tables to create a scenario skeleton → expand only what the session requires → link played scenarios into an emerging campaign web. The same pipeline works for dungeon crawls, corporate runs, wilderness jobs, and political operations.

Seed generation is fast (5 minutes); expansion is proportional to available prep time. Nothing forces more than the minimum.

The Core Seed Structure

Every playable scenario reduces to four elements:

Seed Element	What It Defines	Minimum Viable Form
Client	Who is giving the work and why	Name, domain, one demand
Objective	What the scenario requires the party to accomplish	One verb + one object
Location	Where the scenario plays out	One site type + one detail
Condition	The complicating modifier	One constraint, deadline, or competing interest

A scenario with all four elements is immediately playable. The referee supplies everything else at the table through improvisation seeded by these four inputs.

Fast Mission Pipeline (5 Minutes)

The Running Silent procedure. Usable for any system, not just cyberpunk:

Client: roll on a domain-sorted client table, or choose a known faction from the campaign.
Objective: roll d6 on an objective table. Common objectives: Destroy, Protect, Retrieve, Plant, Investigate, Cover Up, Infiltrate, Escort.
Location: roll on a site-type table + roll on a features/details table. Two rolls, not one.
Condition: roll d20 on a conditions table. Examples: "no loose ends," "time-critical," "rival team also hired," "target is mobile," "one contact in the building."
Pressure level: place the party on the hex-flower matrix or choose a starting tension level (quiet / tense / active).

Combine all five into one sentence: "[Client] wants you to [objective] [location target], but [condition]."

Objective Taxonomy

Eight objective types cover the vast majority of scenario structures. Each implies a different primary challenge type:

Objective	Primary Challenge	Typical Failure Mode
Destroy	Access + execution	Collateral damage, escalation
Protect	Endurance + resource management	Attrition, surprise attack
Retrieve	Infiltration + extraction	Complications on exit
Plant	Stealth + placement	Discovery before activation
Investigate	Exploration + information gathering	Missing key clue; false trail
Cover Up	Speed + elimination of evidence	Witnesses, partial discovery
Infiltrate	Identity + social navigation	Cover blown mid-op
Escort	Movement + threat management	Asset is a liability

Design note: mix objective types across sessions to avoid repetition. A campaign that only uses Retrieve and Destroy narrows mechanical diversity rapidly.

Staged Expansion from Seed to Campaign

Scenarios generated from seeds can remain one-shots or be expanded through three stages:

Stage 1: Single-Site Run

One location, one objective, one session. Seed → site sketch → key elements. No linking required. Keep notes to one page.

Stage 2: Active Complications

Animate the opposition. Add an adversary roster (who is where at start, how they respond to intrusion). Add a secondary complication (a rival party, a time pressure, a hidden agenda from the client). The seed becomes a situation, not a static encounter.

Stage 3: Linked Arc

Once two or more scenarios share a client, faction, or location, link them explicitly. Surface that link through callbacks (NPCs reference previous events), consequence seeds (unresolved scenario outcomes create new jobs), and partial information (each scenario reveals one piece of a larger picture).

An arc does not need to be pre-planned. It emerges when the referee tracks what was resolved and what was left dangling.

Scenario Linking Patterns

These three patterns convert isolated generated seeds into a coherent campaign web:

Sequential: Scenario B follows directly from Scenario A's resolution. One outcome is the setup for the next job. Linear but legible.

Parallel: Two or more scenarios are active simultaneously. Party chooses which to pursue first. Unpursued scenarios advance on their own.

Convergent: Separate scenarios share a hidden connection (same client using different fronts, same location, same MacGuffin). Party discovers the connection through play rather than being told.

The third pattern is the most satisfying but requires the referee to track which scenarios share a common element.

Implementation Variants

Running Silent Fast Pipeline (Cyberpunk)

Domain-sorted client tables (d6 domain → d100 type), structured 8-objective table, location + features tables, conditions table. Full pipeline in 5 rolls. The hex-flower matrix replaces a separate pacing clock. Contact tables integrated: the job source is also a contact acquisition moment.

Perilous Wilds Travel-Move Pipeline (Wilderness)

Scenarios emerge from move outcomes rather than pre-generated tables. Scout a head (10+) → discovery becomes a job. Forage (6−) → danger becomes the scenario. The pipeline is backward: travel generates the seed rather than the seed generating the travel.

Staged Prep Progression (Sci-Fi GM Prep)

Stage 1 (single location) → Stage 2 (active opposition) → Stage 3 (clue investigation) → Stage 4 (linked conspiracy arc). Each stage adds one layer of complexity. The referee advances to the next stage only when the current stage is running smoothly. This makes the pipeline a skills-progression tool, not just a scenario generator.

Design Principles

Seed before expanding. Confirm the four elements are coherent before adding layers. A coherent seed plus good improvisation beats an elaborate script.
Objectives define challenge types. Choose objectives that require different skills and approaches across the session series, not whatever the dice give.
Linking is retroactive. Two isolated seeds become an arc when the referee notices they share an element and makes that connection explicit in play.
Conditions are the complication engine. The fourth seed element is where the scenario becomes interesting. Never skip conditions.

Failure Modes

All seeds, no execution. Generating 10 seeds in prep adds no value if none are run. Generate one or two, run them, link retroactively.
Overloaded conditions. Multiple conditions per scenario fragment focus. One condition per scenario. Additional complications emerge through play.
Linear linking. Connecting scenarios only in a chain produces railroad structure. Introduce at least one parallel or convergent link per three sessions.