Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/grcengineering/companion/reading-guidenpx skills add grcengineering/companion --skill reading-guidegit clone --depth 1 https://github.com/grcengineering/companionWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00056 | $0.00478 |
| Opus 5 | $0.00028 | $0.00239 |
| Sonnet 5 | $0.00011 | $0.00096 |
| Haiku 4.5 | $0.00006 | $0.00048 |
Grade A, and why
reading-guide scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
reading-guide
What
Select the fewest useful readings from the corpus and turn them into an active learning path.
When
- The learner asks what to read.
- The learner wants sources for a concept.
- A concept needs citation grounding.
- The learner has a narrow research or study question.
Not For
- A full multi-week learning path. Use
learning-path-designer. - A first explanation without sources. Use
concept-tutor. - Fabricating citations outside the corpus.
Inputs
- Learner's current question and goal.
knowledge/index.json.- Relevant files under
knowledge/.
Steps
- Clarify the current learning question.
- Search the corpus index for the smallest relevant set.
- Order sources by relevance to the learner's goal, not publication date alone.
- Explain why each source matters.
- Give one active recall question per source.
- Suggest one follow-up practice move.
Validation
- Every recommended source exists in the corpus.
- The list is short enough to complete.
- Each source has a reason and a recall prompt.
Gotchas
- If the corpus lacks a source, say so and label any advice as a general heuristic.
- Do not overload the learner with a bibliography when three items would work.
- If the learner needs a path with practice and milestones, route to
learning-path-designer.
Failure Modes
- Citation theatre: do not cite files that do not support the point.
- Passive reading: every source needs a retrieval question.
- Bad ordering: relevance to the learner's goal outranks recency.
Examples
- User asks "What should I read on evidence?" -> Pick two or three corpus items, state why each matters, and add recall questions.
- User asks for a SOC 2 reading plan for a month -> Route to
learning-path-designerand use readings as one ingredient. - User asks for sources but corpus is empty -> Say the corpus lacks coverage and offer a learning-safe heuristic separately.
What ships with it
3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 63 lines · 56 tokens per session scan A b096e018626f
reading-guide is a skill published in the GitHub repository grcengineering/companion (32 stars, last pushed 3mo ago), licensed MIT. It adds 56 tokens to every session and 478 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
control-mapping
Maps NIST controls to FedRAMP requirements and documents. Use when helping with control implementation, compliance mapping, security baseline alignment, or understanding control requirements.
frmr-analysis
Analyzes FedRAMP FRMR documents to extract control mappings, KSI entries, and version changes. Use when the user asks about FedRAMP requirements, control mappings, compliance data, or needs to understand FRMR document content.
GRC & Compliance
Governance, risk, and compliance — risk assessment and scoring, control mapping across NIST CSF 2.0 / ISO 27001:2022 / SOC 2 / CIS Controls v8, gap analysis, audit evidence preparation, and security policy generation.
mapping-builder
Build a reviewed crosswalk (RequirementMappingSet YAML library + review xlsx/csv) between two CISO Assistant framework YAML files using Claude itself as the reasoning engine. Zero infrastructure — stdlib + pyyaml only, no embedders, no LM Studio, no Qdrant. Use when the user asks to map / crosswalk / generate a…
reference-controls-enricher
Enrich a CISO Assistant framework YAML by linking each assessable requirement to reference control URNs from the central doc-pol library (CISO Assistant Key Reference Controls). Produces a reviewable xlsx and patches the framework YAML in place. Use when the user asks to "add reference controls to framework X", "link…
ciso-assistant-basic-risk-assessment
Guide users through a basic risk assessment workflow in CISO Assistant, from asset identification to scenario creation. Use when: (1) User wants to start a risk assessment from scratch (2) User mentions "risk assessment", "identify risks", "threat scenarios", or "risk register" (3) User asks about qualitative vs…