CtrlK
BlogDocsLog inGet started
Tessl Logo
PluginPublic

ainativedev/aidevcon-2026-nyc

AI Native DevCon 2026 New York — practical agent skills grounded in conference talks

Quality

92%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

Overview
Quality
Evals
Security
Files

context-design-template.mdskills/talk-boeckeler-context-lifecycles/references/

Context delivery design and audit

This is a Tessl working template applying the talk. Its fields, ownership model, tests, and proposed measurements are recommendations, not a schema specified by the speaker. Use only the sections that help the user's decision, and leave unavailable evidence explicitly unknown.

Task and context inventory

State the decision or task, caller, delivery phase, and scope of the material actually inspected.

Information or instructionPurpose for this taskCodified / ambient / mixedWhy this lifecycle fitsSource and version/dateOwner and access scope
To inspectNeeded behavior or knowledgeTo decideReuse and rate of changeEvidence requiredTo confirm

Classify the information rather than assuming every wiki page is ambient or every Markdown file is codified. Split mixed documents if stable instructions and changing facts need different maintenance. Keep essential instructions reliably delivered, and retrieve task-specific evidence when needed.

Lifecycle decisions

Context typeStageDesign decisionEvidence/checkOwner
CodifiedGenerateHow instructions are authored and reviewedVersioned source; intended behaviorTo assign
CodifiedEvaluateRepresentative tasks and edge casesEvidence of correct applicationTo assign
CodifiedDistributeHow the intended version reaches agentsInstalled/delivered versionTo assign
CodifiedObserveHow usage and failures become updatesTraces or user feedbackTo assign
AmbientIngestSources, synchronization, backfill, freshness, identity, permissionsMissed-update and access-boundary checksTo assign
AmbientEvaluateQuestions, gold evidence and answers, edge casesRetrieval coverage and reasoning qualityTo assign
AmbientAssembleCaller/task, relevant sources, conflict handling, budgetSelected evidence, unresolved conflicts, latency/costTo assign
AmbientObserveInput quality and engine qualityDisputed sources, poor-answer clusters, regression casesTo assign

Conflict record

  • Task and question being answered:
  • Conflicting claims, with source locations and dates:
  • Which statements describe current implementation, intended behavior, or historical decisions:
  • Identity/expertise evidence, if relevant:
  • Code or other corroborating evidence inspected:
  • Resolution and reasoning, or unresolved uncertainty:
  • Next verification and responsible owner:

Use code to inspect current behavior, not to automatically overrule a requirement. Do not suppress a real conflict because the answer would be more convenient without it. Retain access boundaries when returning evidence to the caller.

Proposed evaluation cases

For each case record the question, caller/access scope, expected source documents, expected answer or abstention, and observed result. These cases extend the talk's suggestions into a practical starting set:

  1. Retrieve a known relevant decision during planning and explain its relation to the current code.
  2. Resolve a stale-document conflict using trustworthy current evidence, or explicitly report that it remains unresolved.
  3. Avoid treating duplicate or near-duplicate sources as independent confirmation.
  4. Decline to invent an answer when the necessary evidence is absent.
  5. Deliver a reusable review instruction even when related ambient material is sparse.
  6. Handle changed permissions and missed synchronization without returning unauthorized or falsely current information.
  7. Compare retrieval depth, answer quality, latency, and known cost for different caller needs.

Choose project-specific thresholds before interpreting results. The talk does not supply a universal accuracy target, token budget, or synchronization interval. Record what is unmeasured and distinguish correct retrieval from correct reasoning.

Findings and improvement loop

Finding or hypothesisEvidenceConsequenceProposed changeVerification
To inspectSource/trace neededEffect on the taskSmallest useful changeCase that should improve

Separate changes to input knowledge from changes to the context engine. For example, repeated conflicts may indicate a stale document, an overly strong code preference, or a retrieval failure; inspect before choosing a fix. Turn each confirmed failure into a regression case and monitor whether the issue recurs after the change.

skills

talk-boeckeler-context-lifecycles

tile.json