assumptions
SkillDatabases & dataManage the assumption registry — track, verify, and query assumptions about external dependencies and system behavior. Prevents costly rediscovery of known failures.
Available today. Use it from your connected AI after setup.
No other account needed.
Connect ahel once, and every AI you use reads what you have installed.
Then ask your AI: use the assumptions skill
What this skill tells your AI
The instructions your AI receives, as published by kastalien-research/thoughtbox in .agents/skills/assumptions/SKILL.md and read by ahel’s review.
Manage assumptions: $ARGUMENTS
Commands
Parse the first word of $ARGUMENTS to determine the command:
list — Show all tracked assumptions
- Read all
.assumptions/*.jsonlfiles - Parse each line as a JSON record
- Display sorted by confidence (lowest first) or staleness (oldest verification first)
- Format as a table: ID | Category | Claim | Confidence | Last Verified | Status
add — Register a new assumption
Parse remaining arguments for: --category, --claim, --evidence, --source
Create a new assumption record in .assumptions/registry.jsonl:
{
"id": "<category>-<short-slug>",
"category": "<api|dependency|behavior|environment|tooling>",
"claim": "<what we assume to be true>",
"evidence": "<what supports this assumption>",
"source": "<where this was discovered>",
"confidence": 0.8,
"created": "<ISO 8601>",
"last_verified": "<ISO 8601>",
"verification_method": "<how to test this>",
"failure_history": [],
"status": "active",
"blast_radius": "<what breaks if this assumption is wrong>"
}
verify — Re-verify an assumption
- Read the assumption record by ID
- Execute the verification method (may involve web search, API calls, or code checks)
- Update
last_verifiedtimestamp andconfidencescore - If verification fails, add to
failure_historyand reduce confidence - If confidence drops below 0.3, mark status as
suspectand warn
stale — Show assumptions that need re-verification
- Read all assumption records
- Filter to those where
last_verifiedis more than 14 days ago - Sort by blast_radius (highest first)
- Display with suggested verification actions
seed — Seed registry from MEMORY.md gotchas
- Read MEMORY.md
- Extract entries from "Gotchas", "Known Bugs", and "MCP Knowledge API Gotchas" sections
- For each entry, create an assumption record with:
- category: inferred from content (api, dependency, behavior, etc.)
- claim: the gotcha statement
- evidence: "Discovered empirically" + date from MEMORY.md
- confidence: 0.9 (verified by experience)
- verification_method: suggested test
Schema
Each assumption record:
| Field | Type | Required | Description |
|---|---|---|---|
| id | string | yes | Unique identifier (category-slug format) |
| category | enum | yes | api, dependency, behavior, environment, tooling |
| claim | string | yes | What we assume to be true |
| evidence | string | yes | What supports this claim |
| source | string | yes | Where this was discovered (session, test, docs) |
| confidence | float | yes | 0.0 to 1.0 confidence score |
| created | string | yes | ISO 8601 creation timestamp |
| last_verified | string | yes | ISO 8601 last verification timestamp |
| verification_method | string | no | How to re-test this assumption |
| failure_history | array | no | Past verification failures with timestamps and details |
| status | enum | yes | active, suspect, retired, verified |
| blast_radius | string | no | What breaks if this assumption is wrong |
| dependencies | array | no | Other assumptions this depends on |
Output
Always end with a summary:
## Assumption Registry Status
Total: {N} assumptions
Active: {N} | Suspect: {N} | Retired: {N}
Stale (>14 days): {N}
Highest blast radius unverified: {assumption_id}
Signals
- GitHub stars
- 64
- Forks
- 20
- Last commit
- Jul 2026
Advanced
- Catalog kind
- skill
- Gateway key
assumptions- Source
- github.com/kastalien-research/thoughtbox