Product · Scenario injects

MyMuse

Create injects, such as news articles, press releases and social media posts, that can be incorporated into an information scenario.

Contact for pricing Ask a question

MyMuse supports creating injects for information scenarios

MyMuse helps exercise designers, instructors and trainees create realistic injects for live, virtual and constructive exercises: news articles, press releases, reports, think-tank pieces and social media posts for X and Telegram. Injects stay linked to the scenario’s events and personas, and support information missions such as public affairs responses, influence campaigns and countering adversarial messaging.

What you can do with MyMuse

Create injects in minutes, not hours. Produce ten posts and a news article in under an hour, instead of 3 to 12 hours by hand.
Stay consistent with the scenario. Work from the current day’s actors, accounts, narratives and events.
Message across channels. Coordinate one narrative across news, press releases, reports and social media.
Adapt during the exercise. Respond to kinetic events, weather or unexpected developments without rewriting the scenario.
Run real experiments. Replay a scenario with different injects to compare courses of action.

Technical details

Platform	PC
Language	Python
AI	Works with the LLM of your choice
Users	Unlimited simultaneous users; exports merge into one file
Output	AESOP-compatible JSON with images (folder, per-inject folders or embedded)
Requires	An AESOP scenario; injects appear in SynSM-generated data
Scale	From one to several thousand synthetic actors and hundreds of Telegram channels

Why MyMuse

A different kind of exercise

Capability	Traditional exercise	MyMuse exercise
Manual content creation	High	Low
Consistency with scenario	Variable	Automatic
Multi-channel messaging	Labor intensive	Automated
Experimental repeatability	Limited	High
Ground truth availability	Partial	Complete
Adaptive response measurement	Limited	Built-in
Dynamic narrative testing	Difficult	Easy

Time to learn: < 20 min
10 posts + 1 article: < 1 hour
Without MyMuse: 3–12 hours
Simultaneous users: Unlimited

More about MyMuse

How MyMuse works

Details from the MyMuse 2026 overview and Quick User Guide.

Two user modes

Super-user (scenario designer or instructor): full access to every scenario actor, plus controls to change the scenario mid-run.
General user (trainee or support staff): posts from assigned accounts and can create new unattributed accounts.

Injects you can create

Social media posts for X and Telegram: originals, replies, reposts and quotes, with images and mentions
News articles with headline, outlet and URL
Press releases from companies, agencies, NGOs and governments
Long-form reports and think-tank pieces
Custom actors that join the scenario

Change the scenario mid-exercise

Super-users can inject new events, update an actor’s attributes, silence actors, adjust a group’s behavior and push a narrative across channels, so the information environment evolves in real time.

Consistent by design

MyMuse loads each day’s scenario with its actors, platform accounts, narratives, issues and groups. Every inject is tied to a real actor, issue and narrative, so content stays consistent with the storyline. One article expands into posts from the outlet and its reporters, and one press release into an announcement and follow-on coverage.

AI assistance

Draft posts, headlines and article bodies from the actor, platform, topic and stance you set
Generate article and press-release images
Bulk-draft 2–12 posts for an account, grounded in its topics and imported messages
Works with the LLM of your choice; every draft can be edited or discarded

Part of GhostField

MyMuse works within the GhostField system. Each day, users create injects that are exported as a JSON file, merged with other inputs and passed to SynSM, which generates the next day’s information environment, including how the simulated population responds. Pair it with our Information Environment Scenario Creation service.

Built for experimentation

Compare no-inject baselines with treatment runs to measure the effect of specific messages
Test alternative courses of action and messaging strategies
Study narrative competition and second- and third-order effects
See likely top reactors and estimated reach before an inject ships

Ground truth you can trust

Every inject is recorded with its source actor, timing, content, intended narrative and distribution pathway. That makes it possible to validate narrative detection, influence metrics and misinformation tools against known ground truth.

Training and assessment

Instructors can see whether trainees detect emerging narratives, respond appropriately to adversarial messaging and coordinate their messaging, with realistic feedback from a population that adapts to their actions.