CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

recall-before-task

Recall relevant memories from Hindsight before starting a task. Injects past learnings, decisions, and context into the current session.

Quality

81%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

High

Do not use without reviewing

SKILL.md
Quality
Evals
Security

Quality

Content

85%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is a lean, well-structured, and actionable instruction set for a simple, single-purpose skill — clear sequencing with explicit edge-case handling and an exact output template. The only minor improvements are trimming the small redundancy between step 3 and the template, and optionally adding a concrete recall-call example.

DimensionReasoningScore

Conciseness

The body is efficient procedural guidance with no explanations of concepts Claude already knows — it only states what to do with the Hindsight tools. Minor over-explanation keeps it below anchor 5: the four synthesis categories ("Hard rules", "Relevant patterns", "Gotchas", "Related decisions") are defined in step 3 and then repeated verbatim in the output template, a small redundancy that could be tightened.

4 / 5

Actionability

Guidance is concrete and executable for an instruction-only skill: it names the Hindsight `recall` tool, gives three specific query strategies (full description, file/module paths, domain concepts), and provides an exact output format template. It stays below anchor 5 only because there is no example of an actual recall invocation or a filled-in example of the synthesized output; minor gaps, matching anchor 4.

4 / 5

Workflow Clarity

The five steps (parse → recall → synthesize → present → ask user) are clearly sequenced and unambiguous, with explicit edge-case and validation handling: "If no memories are retrieved, say so explicitly" and "Do not fabricate memories. Only present what Hindsight actually returns." The pre-formatted template acts as a checklist for step 4. This is a non-destructive, non-batch skill, so no feedback loop is required; the sequence matches anchor 5's clear-steps-with-explicit-checkpoints standard.

5 / 5

Progressive Disclosure

The skill is under 50 lines, entirely self-contained with no bundle files (no references/, scripts/, or assets/ exist), and no external references are needed. Content is well-organized into "Steps", "Important", and a template block, which per the rubric's simple-skill guideline earns a 5 with just well-organized sections.

5 / 5

Total

18

/

20

Passed

Description

67%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is concrete, specific, and clearly identifies its niche (Hindsight memory recall), with an explicit timing cue for when to use it. Its main weaknesses are limited natural trigger-term coverage and a lack of an explicit "Use when..." clause with concrete trigger phrases.

Suggestions

Add a "Use when..." clause with concrete triggers, e.g., "Use when starting a new task, when the user asks to remember or apply past learnings, or when resuming work on a previously discussed project."

Include natural keyword variations users would actually say, such as "remember", "previous sessions", "past conversations", or "prior work".

Mention the synthesis/presentation step (organizing recalled memories into a context summary) so the capability coverage is comprehensive.

DimensionReasoningScore

Specificity

The description lists several concrete actions — "Recall relevant memories from Hindsight" and "Injects past learnings, decisions, and context" — naming the specific tool and its objects. It sits noticeably above the anchor-3 example (domain plus only 1-2 actions), but minor gaps in coverage (no mention of synthesizing or presenting the recalled context) keep it below the comprehensive anchor 5.

4 / 5

Completeness

The "what" is clear (recall memories from Hindsight and inject past learnings, decisions, and context), and an explicit when-equivalent is stated ("before starting a task"), so the missing-'Use when' cap at 3 does not apply. However, the when lacks concrete trigger phrases and specificity, matching anchor 4 rather than anchor 5's "clearly and explicitly answers both what AND when with concrete trigger phrases".

4 / 5

Trigger Term Quality

Relevant keywords like "memories", "recall", "learnings", "decisions", "context", and "Hindsight" are present, but common natural variations users would say are missing — e.g., "remember", "previous sessions", "past work", "prior conversations". This matches the anchor for having some relevant keywords while missing common variations or synonyms; it is above anchor 2 (only generic keywords) and below anchor 4 (good coverage with just a few terms missing).

3 / 5

Distinctiveness Conflict Risk

Naming the proprietary "Hindsight" tool and the memory-recall niche gives it a distinct identity with minimal overlap risk, fitting anchor 4. It falls short of anchor 5 because the trigger is purely temporal ("before starting a task") rather than topically distinct, which could overlap with general planning or context-gathering skills at task start.

4 / 5

Total

15

/

20

Passed

Validation

100%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository

AndreJorgeLopes/devflow

Path

skills/recall-before-task/SKILL.md

Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.