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/christopherlouet/claude-base/feature-flagsnpx skills add christopherlouet/claude-base --skill feature-flagsgit clone --depth 1 https://github.com/christopherlouet/claude-baseWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/christopherlouet/claude-base/feature-flags)<a href="https://agentmods.dev/skills/christopherlouet/claude-base/feature-flags"><img src="https://agentmods.dev/badge/skills/christopherlouet/claude-base/feature-flags.svg" alt="Measured on agentmods" height="20"></a>What 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 5.1 | $0.00030 | $0.00651 |
| Opus 5 | $0.00015 | $0.00326 |
| Sonnet 5 | $0.00006 | $0.00130 |
| Haiku 4.5 | $0.00003 | $0.00065 |
Grade A, and why
feature-flags 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.
How it starts
The opening of the file, as written. The whole thing — 49 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Feature Flags (pointer)
SDK code, targeting rules syntax and dashboard integration drift on each vendor release and are canonical at:
- LaunchDarkly — launchdarkly.com/docs (SaaS, advanced targeting + experimentation)
- Unleash — docs.getunleash.io (self-hosted open source)
- ConfigCat — configcat.com/docs (SaaS, generous free tier)
- PostHog Feature Flags — posthog.com/docs/feature-flags (paired with product analytics; see also recipe)
- OpenFeature — openfeature.dev (vendor-neutral SDK standard — use to avoid lock-in)
Use-case taxonomy (version-agnostic)
| Type | Purpose | Lifetime |
|---|---|---|
| Release toggle | Deploy inactive code; flip on after smoke test | Short (days/weeks) |
| Experiment toggle | A/B test; emits exposure events for analysis | Bounded by experiment duration |
| Ops toggle | Circuit breakers, kill switches, degraded modes | Long-lived |
| Permission toggle | Feature gating by role/plan/cohort | Permanent (treat like config) |
Choose by lifetime: short → cheap implementation, long-lived → invest in observability + naming discipline.
Foundation discipline (keep across releases)
- Default OFF: every flag defaults to its conservative value (usually off). A flag that ships defaulting to ON is a hidden behaviour change.
- 2-sprint rule: remove release toggles within 2 sprints of full rollout. Stale flags accrue as tech debt — surface them via
qa-tech-debt. - Log every evaluation: missing evaluation logs make debugging "why did user X see variant Y" impossible. Vendor SDKs offer this; if rolling custom, log it.
- No business logic in flag values: a flag is a boolean (or enum); complex conditions belong in code paths the flag selects, not inside the flag service.
- Naming convention:
<scope>_<feature>_<variant>(e.g.checkout_express_enabled). Scope-first sorts/filters cleanly in dashboards.
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 · 49 lines · 30 tokens per session scan A 08d11e513a94
feature-flags is a skill published in the GitHub repository christopherlouet/claude-base (5 stars, last pushed yesterday), licensed MIT. It adds 30 tokens to every session and 651 once invoked, about $0.0002 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 2026-09-03.
Other skills, from other repositories
churn-prevention
../../../marketing-skill/skills/churn-prevention/SKILL.md.
content-production
../../../marketing-skill/skills/content-production/SKILL.md.
board-meeting
../../../c-level-advisor/skills/board-meeting/SKILL.md.
boardroom
../../../c-level-agents/skills/boardroom/SKILL.md.
cpo-review
../../../c-level-agents/skills/cpo-review/SKILL.md.
moai-kanban-foreman
One unattended kanban foreman iteration: watch the backlog queue, dispatch the next operator-picked card to an isolated worker, collect completion evidence on read (not on claims), and report. This is the body the project's loop.md driver invokes each iteration of a bare /loop; it can also be invoked directly to test…