Check sources
Skill control-decks/reality-check/plugins/reality-check/skills/check-sources
A verification deck for checking sources, assumptions, and concrete targets without replacing the result.
npx -y skills add control-decks/reality-check --skill check-sourcesAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- 17 days oldThe repository was created 17 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Verify the sources supporting the current HACP Working Object, distinguish direct evidence from inference and unknowns, and append a sources-checked annotation without replacing the object. Use only when the user invokes CHECK SOURCES or explicitly asks to verify the current sources.
SKILL.md
2.1 KB, 422 tokens by cl100k_base, as published. Nobody here has run it
π Check Sources
ID: reality-check/check-sources
HACP: 0.4
Kind: operation
Mode: annotate
Traits: read-only, evidence-annotation, source-verification
Default Binding: Claims and sources in the current object or request
Accepts: hacp/content, hacp/result
Produces: hacp/result, preserving the input kind and content and adding
reality-check/sources-checked with traits evidence and source-evidence
Duration: once
Effect: Verify material claims against the best allowed sources and append a scoped evidence annotation without replacing the current object.
Check
- Identify only the claims whose truth matters to the object's use.
- Inspect the cited or best allowed primary sources. Revalidate time-sensitive claims.
- Mark each material claim
supported,partial,contradicted,unverified, orinference. - Append source identity, freshness, access limitations, resolved Binding, traits, and source. Preserve the object's Binding, status, kind, content, and existing annotations.
Honor active controls before access. When external access is blocked, consult
only local files, cached artifacts, command output, and conversation context;
mark remote claims unverified locally. Never name or depend on the card that
established the control.
Limits
- A plausible statement is not evidence.
- A secondary source cannot silently stand in for an available primary source.
- Never fabricate citations, quotes, dates, or tool observations.
Format
Contribute β π **CHECK SOURCES** to the one complete combo trace, or begin
with > π― **<binding>** β π **CHECK SOURCES** when used alone. Then use
Sources for the material classifications. Never emit a second trace or
deck-status line.
What ships with it: 2 files
18.9 KB alongside SKILL.md
agents/
- openai.yaml288 B
assets/
- icon.png18.6 KB
Gives 0 of the 12 instructions most quality gates skills give in 422 tokens
Counted across 1,195 of the 2,094 authors here whose files we hold, read 2026-08-07
- Read the output and check the exit codein 54 of 1195, across 14 files
- Verify requirements using a line-by-line checklistin 53 of 1195, across 12 files
- Identify the verification command proving the claimin 51 of 1195, across 12 files
- Run the full verification commandin 50 of 1195, across 11 files
- Verify output confirms the claimin 49 of 1195, across 12 files
- Check version control diff after agent delegationin 46 of 1195, across 6 files
- State claim with evidencein 44 of 1195, across 4 files
- Run the test suitein 33 of 1195, across 26 files
- Keep state in memory by defaultin 27 of 1195, across 6 files
- Make prototype runnable with one commandin 26 of 1195, across 5 files
- Produce a verification reportin 25 of 1195, across 14 files
- Detect the package manager from lockfilesin 24 of 1195, across 5 files
Said here and by no other author read
- identify only material claims
- inspect primary sources
- revalidate time-sensitive claims
- mark each material claim status
- append source identity and freshness
- preserve object binding and existing annotations
Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.