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 agents/paulbreuler/limps/001-context-resolvergit clone --depth 1 https://github.com/paulbreuler/limpsWhat 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 | $0.00000 | $0.00500 |
| Opus 5 | $0.00000 | $0.00250 |
| Sonnet 5 | $0.00000 | $0.00100 |
| Haiku 4.5 | $0.00000 | $0.00050 |
Grade A, and why
001-context-resolver 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
Agent 001: Context Resolver
Objective
Build the context inheritance engine that resolves cascading context from workspace → project → plan → agent.
Tasks
-
Context layer loading (
src/context/resolver.ts)- Load workspace layer (priority 0)
- Load project layer if exists (priority 10)
- Load plan layer (priority 20)
- Load agent layer (priority 30)
- Track source file for each value
-
Merge strategy (
src/context/merge.ts)- Higher priority wins for scalar values
- Lists concatenate (child adds to parent)
- Objects deep merge
override: truereplaces entirelyinherit: falseopts out
-
Resolve CLI (
src/cli/commands/context/resolve.ts)limps context resolve— workspace onlylimps context resolve <plan>— with plan contextlimps context resolve <plan> <agent>— full stack--diffshows what each layer contributed--jsonfor machine consumption
Resolution Algorithm
async function resolveContext(planId?: string, agentId?: string): Promise<ResolvedContext> {
const layers: ContextLayer[] = [];
// Build layer stack
layers.push(...await loadWorkspaceLayers());
if (planId) {
const projectId = await findProjectForPlan(planId);
if (projectId) layers.push(await loadProjectLayer(projectId));
layers.push(await loadPlanLayer(planId));
if (agentId) layers.push(await loadAgentLayer(planId, agentId));
}
// Merge with priority resolution
return mergeLayers(layers);
}
Acceptance Criteria
-
limps context resolveoutputs workspace context -
limps context resolve 0042 001includes all layers - Higher priority values override lower
-
--diffshows contribution per layer - Resolution completes in < 100ms
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 · 69 lines · 0 tokens per session scan A 65029bc585f7
001-context-resolver is an agent published in the GitHub repository paulbreuler/limps (10 stars, last pushed 6mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 500 tokens. 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 2026-08-31.
Other agents, from other repositories
monitor
Reviews code for correctness, standards, security, and testability (MAP).
reflector
Extracts structured lessons from successes and failures.
spirit-check
Fresh-context reviewer that checks whether a SPEC's interpretation matches the user's verbatim ask. Reads ONLY the user's literal words + the SPEC's Behavior section — nothing else. Flags interpretation drift (e.g., user said "function", SPEC describes "deployed service"). Dispatch it after writing a SPEC, before it…
ai-layer-reviewer
Fresh-context adversarial reviewer for a freshly-authored .claude/ artifact (skill, hook, or agent). Dispatched by /author's design-review gate after fill, before verify. Attacks the artifact's DESIGN — wrong step graph, missing gate, prose describing a deterministic step, a description that will never fire, unhandled…
rule-efficacy-reviewer
Fresh-context reviewer that checks whether compressing a rule dropped any behavior. Dispatched by /author's efficacy-gate step during rule compression. Receives the original rule text, the compressed rewrite, and the list of dropped clauses — blind to the author's rationale — and returns a per-clause verdict of…
discovery-analyst
Use proactively during /fp:init to perform Phase 1 (Discovery) of the first-plan plugin. Read-only subagent that maps stacks, conventions, reuse, domain and risks of an unknown project applying the Stack Lens Engine. Returns structured findings to be written to .first-plan/. Do NOT use for execution or modifications …